id stringlengths 9 34 | models listlengths 4 4 | completions listlengths 4 4 | gpt_inference stringlengths 4 4.72k | instruction stringlengths 13 1.38k | scores listlengths 4 4 | score_variance float64 0 0.25 | correlation float64 -0.84 1 | average_score float64 -0.02 1 |
|---|---|---|---|---|---|---|---|---|
svit-complex_reasoning-5530 | [
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b",
"instructblip-vicuna-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The response is moderately helpful as it provides a specific suggestion to place the coffee maker on the countertop and considers cleanliness and space issues. However, it also goes into detail about organizing cabinets and d... | To make the coffee maker more accessible in this kitchen layout, consider the following changes:
1. **Relocate the Coffee Maker:**
- Place the coffee maker closer to the sink where water is easily accessible for refilling the water tank.
- Consider positioning it on an open countertop away from corners or obstacles, which will make it easier to use and clean.
2. **Create a Dedicated Coffee Station:**
- Designate a section of the counter specifically for coffee prep. Include the coffee maker, mugs, and coffee supplies in one area.
- Move any unrelated items (e.g., paper towel holder) to another counter to free up space near the coffee maker.
3. **Adjust Electrical Location:**
- If needed, move the coffee maker closer to a more central power outlet or install a new outlet in a convenient spot that minimizes cord crossing across the counter.
4. **Avoid Tight Corners:**
- The existing placement of the coffee maker is in the corner, which may make it feel cramped. Moving it to a more open counter space improves accessibility.
5. **Increase Counter Space:**
- If counter space is limited, consider wall-mounted shelves or under-cabinet storage for other items, freeing up room for the coffee maker.
By effectively reorganizing or dedicating a section exclusively as a coffee corner, you’ll enhance both functionality and accessibility. | How can the layout of the kitchen be improved to make the coffee maker more accessible? | [
0.8268907070159912,
0.846511721611023,
0.8592118620872498,
0.8119560480117798
] | 0.000328 | 0.991961 | 0.836143 |
LLaVAMed-000000004702 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This text provides an inaccurate interpretation of the CT scan image, falsely identifying a large piece of metal connected to the chest wall. The image does not show any metallic objects, so this response is not helpful to th... | This CT scan demonstrates a subdural hematoma, which appears as a crescent-shaped, hyperdense (bright) collection along the inner surface of the skull. Subdural hematomas are caused by the accumulation of blood between the dura mater and the arachnoid membrane, often due to trauma. The shifted midline structures also indicate mass effect, suggesting significant pressure from the hematoma.
Further evaluation and prompt medical intervention are essential if this is a recent finding. | What is the main finding in the CT scan? | [
0.408779501914978,
0.2509617209434509,
0.4225631356239319,
0.35765179991722107
] | 0.004547 | 0.961561 | 0.359989 |
svit-detail_description-903 | [
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b",
"fuyu-8b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "Text 4 does not address the instruction effectively. It lists a series of unrelated items and colors that are not visible in the image, which does not contribute to the user's understanding of the specific image content."
... | The image shows a person seated comfortably in a tan leather armchair. The individual is wearing a red top and blue jeans, with white shoes, and appears relaxed. The room has a warm and cozy environment with visible furniture and decor:
1. **Background Details:**
- A wooden cabinet with glass doors is visible, containing decorative or dining items.
- The wallpaper appears to have a floral or patterned design.
- A chess set is seen on a small table or shelf in the background.
2. **Other Furniture and Items:**
- A wooden chair is partially visible to the left.
- A small side table with a white lamp and books is next to the armchair.
- A red throw or blanket is draped over the armchair.
3. **Flooring and Rugs:**
- The flooring is wooden, with a small woven rug partly visible.
The setting overall gives a homely, lived-in ambiance. | What are the specifics visible in the image? | [
0.4555274248123169,
0.3285968005657196,
0.3099215626716614,
0.24031902849674225
] | 0.006038 | 0.906809 | 0.333591 |
LLaVAMed-000000000530 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"MMICL-Vicuna-13B",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "Text 3 misinterprets the image as being related to learning a new skill or concept and provides an explanation that is unrelated to the actual content, which is about a behavioral conditioning experiment."
},
"eth... | The **conditioning phase** (as shown in Phase 1 of the image) is used to establish associations between specific stimuli and particular outcomes. For instance:
- The aversive stimulus (e.g., the image of the aggressive dog) is paired with a specific shape (e.g., a purple rectangle).
- The neutral stimulus (e.g., the pocket watch) is paired with a different shape (e.g., a green oval).
### Purpose:
The goal of this phase is to "train" participants to associate certain cues (shapes) with unpleasant or neutral outcomes. This preparatory phase serves to:
1. **Condition** the brain to create predictive associations between stimuli and outcomes.
2. Prepare participants for the subsequent **avoidance learning phase**, where they will use these associations to make decisions about avoiding aversive outcomes (as shown in Phase 2, during fMRI scanning).
By conditioning responses beforehand, researchers can study learning, decision-making, and avoidance behavior, as well as neural activity in regions associated with such processes. | What is the purpose of the conditioning phase? | [
0.7575317621231079,
0.6391791105270386,
0.4781074523925781,
0.6769741177558899
] | 0.010343 | 0.98753 | 0.637948 |
LLaVAMed-000000003315 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"instructblip-vicuna-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "Text 4 inaccurately describes the image, misidentifying an air bubble and blood flow where there are none indicated by the arrows. It provides advice that may confuse the user without offering any specific or accurate insight... | In the provided chest X-ray, the **red arrow** and **yellow arrow** point to specific areas of interest within the lungs:
1. **Red Arrow**: Likely indicates an abnormality, such as a localized lesion, opacity, or area of consolidation. This might signify pneumonia, a mass, or another lung pathology.
2. **Yellow Arrow**: Appears to indicate a different region of potential concern. It could represent another abnormal finding, such as fluid accumulation, a separate lesion, or another pathological feature.
Both arrows draw attention to regions that require further clinical evaluation and correlation with the patient’s symptoms, history, and possibly other diagnostic tests. A radiologist or medical professional would determine the exact significance. | What do the arrows indicate in the image? | [
0.634301483631134,
0.5850486755371094,
0.7158727049827576,
0.8330302238464355
] | 0.008807 | 0.959353 | 0.692063 |
LLaVAR-000000005022 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 4 mentions \"quality education\" which is related to the icon, but it does not clearly identify the icon as part of the Sustainable Development Goals, nor does it provide the specific context or significance associated w... | This image represents **Sustainable Development Goal 4 (SDG 4)**, which focuses on **Quality Education**. It is one of the 17 goals established by the United Nations as part of the 2030 Agenda for Sustainable Development.
### **Significance of SDG 4:**
The goal aims to:
- **Ensure inclusive and equitable quality education** for all individuals, regardless of their background.
- Promote **lifelong learning opportunities**.
- Address challenges like lack of access to education, gender inequality in education, and poor quality of learning environments.
**Key Targets of SDG 4:**
- Universal primary and secondary education.
- Equal access to affordable technical, vocational, and higher education.
- Gender equality in education.
- Literacy and numeracy improvement.
- Safe, inclusive learning environments.
The **book and pencil icon** symbolizes learning and education, while the red background adds emphasis to its critical importance. | What is the significance of the white icon with a book and pencil on a red background? | [
0.7834240198135376,
0.856036365032196,
0.9175578355789185,
0.9035390615463257
] | 0.002746 | 0.981075 | 0.865139 |
PMC-VQA-255 | [
"LLaVA-RLHF-13b-v1.5-336",
"LLaVA-RLHF-7b-v1.5-224",
"fuyu-8b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "This response provides a general explanation of MRI planes and their benefits but does not specifically confirm the plane visible in the provided image. It offers some relevant information regarding the positioning of MRI sca... | This breast MRI image was performed in the **sagittal plane**. The image shows a vertical slice from the side of the breast, which corresponds to the sagittal orientation commonly used in breast imaging to provide detailed views of the breast tissue. | What plane was the breast MRI performed in? | [
0.7600061893463135,
0.5906200408935547,
0.056724220514297485,
0.8541011810302734
] | 0.095152 | 0.92825 | 0.565363 |
LLaVAMed-000000003684 | [
"LLaVA-RLHF-13b-v1.5-336",
"LLaVA-RLHF-7b-v1.5-224",
"fuyu-8b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "While the text correctly identifies the organ as lungs, it inaccurately describes the condition of the organ, suggesting decay and presence of maggots, which is not discernible from the image. It adds speculative context not ... | The histological images shown are of the **liver**. The evidence lies in the presence of central veins (CV) and other characteristic features of liver tissue, such as ballooning and lipid inclusions, which are typical of hepatic cells and pathological liver specimens. | What organ is shown in the image? | [
0.382988840341568,
0.3316112458705902,
0.2967323660850525,
0.44279593229293823
] | 0.003036 | 0.993174 | 0.363532 |
LLaVAR-000000008770 | [
"fuyu-8b",
"instructblip-vicuna-13b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 1 simply repeats the first part of the quote without providing any additional context or interpretation regarding its meaning, which lacks significant helpfulness."
},
"ethical_considerations": {
"Rat... | The main message of the quote is about self-improvement and self-worth. It encourages you to continually work on bettering yourself so that you bring value to your relationships and interactions. By doing so, you will attract meaningful connections where you receive wholehearted effort and reciprocity from others. | What is the main message of the quote in the image? | [
0.14478135108947754,
0.4864056706428528,
0.4711030423641205,
0.5197010040283203
] | 0.022966 | 0.950585 | 0.405498 |
svit-conversation-978 | [
"LLaVA-RLHF-7b-v1.5-224",
"fuyu-8b",
"instructblip-vicuna-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The response identifies that the person is laying down and mentions the room is messy, which partially addresses the prompt. However, it inaccurately describes the person's position and lacks detail about the surroundings, wh... | The person in the image is surfing on a wave while riding a surfboard. They are wearing a wetsuit and appear to be balancing skillfully on the water. | What is the person doing in the image? | [
0.25071603059768677,
0.33775511384010315,
0.258208692073822,
0.31797847151756287
] | 0.001403 | 0.95552 | 0.291165 |
llava-detail-291 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response is not helpful because it inaccurately describes the woman as using her phone to take photos, which is not visible or suggested in the image. The mention of people enjoying a rain-free atmosphere contradicts the... | The person in the scene is looking through a refrigerator and appears to be selecting or examining an item—possibly food in a container. The individual is also wearing a headset, suggesting they might be multitasking, such as talking on the phone while organizing or retrieving food. They are holding a plastic bag, possibly to gather or store items. | What's happening in the scene? | [
0.1966724991798401,
0.3277566134929657,
0.24037952721118927,
0.3162502944469452
] | 0.002932 | 0.985206 | 0.270265 |
svit-conversation-3778 | [
"fuyu-8b",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"idefics-9b-instruct"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "Text 2 identifies the main feature as a \"large blue boat,\" which is an accurate description of the image's most prominent element. However, it doesn't acknowledge the red part of the boat or provide context about its surrou... | The main feature in the image is a ship with the name "Paper Moon" docked at a port. It is a cargo vessel, and there are other ships, industrial structures, and residential buildings in the background. The waterfront scenery also includes small kayaks near the ship. | What's the main feature in the image? | [
0.47782665491104126,
0.5990833640098572,
0.6298757791519165,
0.5620877742767334
] | 0.00324 | 0.993545 | 0.567218 |
llava-conversation-3446 | [
"fuyu-8b",
"instructblip-vicuna-13b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Similar to Text 2, this response directly states \"Vanilla\" as the flavor, but it does not provide any context or explanation, which makes the response minimally helpful and potentially misleading."
},
"ethical_c... | This is most likely a slice of vanilla or white cake with white frosting, based on its appearance. Vanilla is a common flavor for cakes with this style, but the exact flavor cannot be determined solely from the image. | What is the main flavor of the cake in the image? | [
0.341055691242218,
0.7085311412811279,
0.7085311412811279,
0.8556883335113525
] | 0.03614 | 0.975313 | 0.653452 |
LRVInstruction-000000000087 | [
"llava1.5-7b",
"llava1.5-7b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The text claims to confirm the presence of a ghostly figure as per the instruction, but there is no such figure in the image provided. Therefore, the response is not helpful because it confirms something that isn't true."
... | This striking artwork showcases a clock framed by intricate sculptures and a decorative archway, which is rich in detail. Upon examining the image, there doesn't appear to be a ghostly figure in the reflection on the clock or near the left edge. If you believe this area holds supernatural intrigue, it may be more about artistic interpretation or the way certain details are perceived. | In the clock's reflections on the archway, can you spot a ghostly figure standing just beyond the left edge? This supernatural element adds intrigue to the artwork. | [
0.886135995388031,
0.8686367273330688,
0.5669375658035278,
0.5514985918998718
] | 0.025376 | 0.696993 | 0.718302 |
svit-detail_description-1132 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Text 1 provides a rich and detailed description of the scene and the skateboarder's activity, mentioning the environment and the contrast created by the lighting, which adds to the comprehension of the image. It doesn't quite... | This image captures a dynamic and playful moment indoors. A man is holding what appears to be a game controller, likely from a Nintendo Wii console, as he engages in an energetic activity, possibly a video game. His posture, with one arm raised and one leg lifted, suggests enthusiasm and physical involvement, showcasing the active, motion-driven gameplay typical of the Wii.
The background features a cozy living room with some furniture visible, including a couch and a soft light fixture hanging from the ceiling. A picturesque beach-themed wall poster adds a touch of decor. In the foreground, everyday items, including a snack and a drink, are visible on a table, further reinforcing the casual, fun setting.
The photo has an artistic feel, with central elements in focus while the edges are slightly blurred. This effect directs attention to the man and his lively movement, emphasizing the action and energy of the moment. The image suggests a lighthearted, social atmosphere, possibly during a friendly gaming session. | Provide a detailed commentary on the image. | [
0.41652002930641174,
0.5140666365623474,
0.1297309696674347,
0.5030554533004761
] | 0.024154 | 0.955694 | 0.390843 |
svit-conversation-2196 | [
"llava1.5-7b",
"fuyu-8b",
"instructblip-vicuna-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The response is very brief and offers minimal information about the types of buildings, only mentioning their height and building material. There is a lack of detailed insight, which minimally addresses the instruction."
... | The image shows two distinct types of buildings:
1. **Historic or older-style building** (on the left): This building features arched windows, detailed ornamentation around the windows, and a classic architectural design, suggesting an older, possibly historic structure.
2. **Modern building** (on the right): This one has a simpler design with clean lines, rectangular windows, and minimal decoration, indicative of more contemporary construction.
Additionally, rooftop structures, such as water towers, can be seen, a common feature in urban areas. | What types of buildings are in the image? | [
0.7085579037666321,
0.5724078416824341,
0.78069007396698,
0.9159281849861145
] | 0.015401 | 0.999987 | 0.744396 |
llava-conversation-4165 | [
"VisualGLM-6b",
"fuyu-8b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The text is helpful because it confirms the presence of one couch and adds contextual information about the living room setup, which could enhance the user's understanding."
},
"ethical_considerations": {
... | There are two couches in the living room visible in the image: one large couch on the left and a smaller one (a loveseat) against the wall. | How many couches are there in the living room? | [
0.7740401029586792,
0.11732999235391617,
0.8614420294761658,
0.8005515933036804
] | 0.091488 | 0.90159 | 0.638341 |
svit-complex_reasoning-3638 | [
"llava1.5-7b",
"llava1.5-7b",
"instructblip-vicuna-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "This text is helpful, it directly addresses the user's question by confirming that the activity is skateboarding based on the clear visual cue of people riding skateboards down the sidewalk."
},
"ethical_considera... | The people in the image appear to be skateboarding or standing on skateboards, likely engaging in a leisurely or social activity. | What seems to be the activity that the people in the image are involved in? | [
0.7907699942588806,
0.7907699942588806,
0.789559006690979,
0.7650281190872192
] | 0.000121 | 0.998071 | 0.784032 |
LRVInstruction-000000009164 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The text is not helpful because it inaccurately describes the hairdryer as purple and being located on a white towel, which is not present in the image. The additional details about the coffee maker and mirror are accurate bu... | There is no purple hairdryer in this image, but there is a white hairdryer attached to the bathroom wall. | Find the purple hairdryer attached to the bathroom wall. | [
0.7542504668235779,
0.8094999194145203,
0.06762748956680298,
0.9679050445556641
] | 0.119132 | 0.893345 | 0.649821 |
svit-complex_reasoning-693 | [
"VisualGLM-6b",
"fuyu-8b",
"instructblip-vicuna-7b",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "Text 2 is highly helpful as it not only states that the image was taken during the day but also narrows down the time to morning or early afternoon based on the behavior of the sheep and the environmental context, offering a ... | This image shows sheep grazing calmly in an open field, which doesn't indicate any specific time of day. However, the lighting suggests partially diffused sunlight with a mix of blue sky and clouds, likely captured during daytime, either in the late morning or early afternoon. The animals' relaxed behavior also indicates a time when it's neither too hot nor too cold, which often aligns with mid-morning to mid-afternoon hours. | Based on the behavior of the animals in the image, what time of day do you think the image was taken? | [
0.8569185733795166,
0.5249452590942383,
0.8266962170600891,
0.754668116569519
] | 0.016912 | 0.998137 | 0.740807 |
svit-complex_reasoning-7989 | [
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b",
"instructblip-vicuna-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "Text 4 provides a correct, albeit brief, analysis by mentioning surfboards and people's movement towards the ocean. However, it introduces ambiguity by suggesting that people might have finished surfing, which is not clearly ... | The image shows several visible indications that suggest people are preparing to surf or are already surfing:
1. **Surfboards**: There are individuals holding and carrying surfboards, a clear indicator of surfing activity.
2. **Wetsuits**: Some individuals are wearing wetsuits, which help with temperature regulation and protection while surfing.
3. **Beach Setting**: The presence of waves and a sandy shoreline, combined with the gear, points to surfing as the primary activity.
4. **Entering the Water**: Some individuals are seen heading into the water with their boards, a clear step in preparing to surf.
5. **Watching the Waves**: The person on the shore might be observing wave patterns to assess conditions before entering.
These visible signs suggest these individuals are either preparing to surf, learning, or already engaging in the activity. | What are some visible indications that suggest the people are preparing to surf? | [
0.7657313346862793,
0.9053696393966675,
0.9271661043167114,
0.8598908185958862
] | 0.003843 | 0.999602 | 0.864539 |
svit-conversation-1699 | [
"LLaVA-RLHF-13b-v1.5-336",
"LLaVA-RLHF-7b-v1.5-224",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "This response is similar to Text 1, giving a general description of the scene with an emphasis on the activities occurring. It offers context about the park and its uses, which is somewhat helpful, but it does not provide an ... | There are at least 11 people visible in the park in this image. | How many people are in the park? | [
0.6067317128181458,
0.7950764298439026,
0.665629506111145,
0.8889122009277344
] | 0.012124 | 0.994376 | 0.739087 |
svit-complex_reasoning-1626 | [
"LLaVA-RLHF-13b-v1.5-336",
"idefics-9b-instruct",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "The response provides a comprehensive and detailed list of potential dangers that a kitesurfer may face, directly addressing the user's query. It includes environmental factors, equipment considerations, and external risks, e... | A possible danger to the kitesurfer in the image could be the proximity to the rocky shoreline. If the kitesurfer loses control or the wind suddenly shifts, they might collide with the rocks, leading to potential injuries or damage to their equipment. Maintaining a safe distance from such hazards is essential in kitesurfing. | What could be a possible danger to the kitesurfer in the image? | [
0.9052419662475586,
0.9381468892097473,
0.9031985998153687,
0.8621664047241211
] | 0.000726 | 0.990713 | 0.902188 |
svit-complex_reasoning-2165 | [
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The response is overly verbose and repetitive, providing an exhaustive list of what the object is not and other irrelevant details, which do not contribute to addressing the user's question effectively."
},
"ethic... | The remote control is located on top of a book that is placed on a table. | Where is the remote control located? | [
0.3775908350944519,
0.9403526782989502,
0.8259828090667725,
0.9186846017837524
] | 0.052043 | 0.994299 | 0.765653 |
LRVInstruction-000000000998 | [
"LLaVA-RLHF-7b-v1.5-224",
"idefics-9b-instruct",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The response describes elements that are not present in the image, which does not help the user with their request. There is no purple kettle or indication of tea-making in the image provided."
},
"ethical_conside... | I don't see a purple kettle boiling on the stove in this image. The stove appears to be clear of any kettle. If you're searching for a specific kettle, it might not be in this picture. | Search for the purple kettle boiling on the stove; it seems like someone's about to make some tea. Can you find the steaming kettle? | [
0.7107683420181274,
0.7592917680740356,
0.8056011199951172,
0.7468017935752869
] | 0.00115 | 0.937826 | 0.755616 |
svit-conversation-1873 | [
"LLaVA-RLHF-13b-v1.5-336",
"fuyu-8b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "This response is very helpful as it not only identifies the game and the players involved but also provides additional context regarding the nature of a baseball game, the roles, and responsibilities of the players, and the c... | This image shows a baseball game. The batter is swinging at a pitch while the catcher is crouched behind home plate. Baseball is a team sport where players attempt to score runs by hitting the ball and running through a series of bases. | What kind of game is happening in this image? | [
0.8564636707305908,
0.8212752342224121,
0.8504782915115356,
0.8344753980636597
] | 0.00019 | 0.995635 | 0.840673 |
svit-conversation-2166 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"fuyu-8b",
"instructblip-vicuna-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The text provides a description of the scene with some context, suggesting the activity they are engaged in, which might be helpful for understanding the image. However, it does not directly answer the specific question of ho... | There are approximately 9 people visible in the image. | How many people are there in the image? | [
0.4232868254184723,
0.5446945428848267,
0.3002566993236542,
0.7530629634857178
] | 0.027927 | 0.948049 | 0.505325 |
llava-conversation-2231 | [
"llava1.5-7b",
"fuyu-8b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Similar to Text 1, this response correctly identifies the activity as skiing and includes the detail of a snow-covered slope, which is visible in the image."
},
"ethical_considerations": {
"Rating": "5",
... | The person in the image is skiing. You can see they are using ski poles, which are commonly used in skiing, and they appear to have two skis on their feet. | Is the person skiing or snowboarding in the image? | [
0.7302266359329224,
0.5849289298057556,
0.7230806350708008,
0.785477876663208
] | 0.005463 | 0.999533 | 0.705929 |
svit-complex_reasoning-9115 | [
"VisualGLM-6b",
"fuyu-8b",
"idefics-9b-instruct",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Text 2 is helpful because it gives a more detailed description of the toppings, suggesting a type of meat (ham), and acknowledges the presence of vegetables. The additional context about it being a traditional Italian-style p... | This pizza appears to have the following toppings:
1. **Tomato sauce** as the base.
2. **Mozzarella cheese** in small melted portions.
3. **Ground meat or sausage crumbles**, likely seasoned.
4. **Herbs or pesto**, adding green spots for added flavor.
It looks like a flavorful and hearty pizza! | What different kinds of toppings can be seen on the pizza? | [
0.8378032445907593,
0.8031454682350159,
0.8307187557220459,
0.7864987850189209
] | 0.00043 | 0.99598 | 0.814542 |
LLaVAMed-000000000919 | [
"LLaVA-RLHF-13b-v1.5-336",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response inaccurately describes the image as depicting a lumbar puncture procedure, which is not present in the MRI scans. The image does not show a patient undergoing a surgical procedure but rather the scans themselves... | The image depicts spinal MRI scans showing a large spinal tumor compressing the spinal cord. It appears to be from the thoracic spine region. Surgery for this condition would likely involve **spinal tumor resection**, which is performed to remove the tumor and relieve the pressure on the spinal cord. Depending on the tumor's type and location, the procedure could involve laminectomy, laminoplasty, or other techniques to access the tumor. | What type of surgery is shown in the image? | [
0.5026658773422241,
0.5693584680557251,
0.6342029571533203,
0.7765922546386719
] | 0.010263 | 0.972756 | 0.620705 |
svit-complex_reasoning-7001 | [
"llava1.5-7b",
"llava1.5-7b",
"fuyu-8b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Text 1 offers a comprehensive explanation of the potential reasons for the paint drips, considering both accidental and intentional causes. It discusses the possibility of the vases being a work in progress or intentionally d... | The paint drips on the vases appear to be an intentional design choice, likely employed as part of an artistic or experimental approach to decorating.
Here are some potential reasons for the drips:
1. **Contemporary Artistic Style**: The drips often reflect a modern or abstract aesthetic, emphasizing imperfections and spontaneity in artistic expression.
2. **Technique Used**: The effect could have been achieved by pouring or dripping paint onto the vases from above, allowing gravity to create the patterns.
3. **Creative Experimentation**: An intent to break conventional norms of smooth and uniform coloring to evoke a sense of fluidity, motion, or unpredictability.
4. **Highlighting Contrast**: The dripped paint motif emphasizes contrast between the base color and the overlaid color, creating a visually striking appearance.
5. **Metaphor or Theme**: The drips might symbolize natural processes (like rain or melting) or convey a specific thematic idea such as freedom or imperfection.
This design adds an eye-catching, tactile, and dynamic element to the vases. | What could be the potential reason for the paint drips on the vases? | [
0.9353344440460205,
0.9482424259185791,
0.8863430619239807,
0.9457868337631226
] | 0.000628 | 0.989106 | 0.928927 |
LRVInstruction-000000001534 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 1 provides some relevant context about the setting but incorrectly identifies the presence of a blue vase. It offers additional details about the room and light source, which, while related to the scene, do not address t... | No, I don't see a blue vase on the countertop in this image. There is a glass upside-down, a lamp, and a container, but there is no blue vase visible. | Can you see a blue vase on the countertop? | [
0.736736536026001,
0.9214993715286255,
0.8402470350265503,
0.9332317113876343
] | 0.006178 | 0.960376 | 0.857929 |
LRVInstruction-000000009009 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-13b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "The response accurately describes the presence of short trees in the background and provides additional context regarding their impact on the scene and the atmosphere they create, directly addressing the user's question."
... | In the background, most of the trees visible are tall, such as palm trees. There don’t appear to be any noticeably short trees in the scene. | Are there any short trees in the background? | [
0.649596631526947,
0.005529951304197311,
0.7374702095985413,
0.8034189939498901
] | 0.101432 | 0.917107 | 0.549004 |
LLaVAR-000000009184 | [
"fuyu-8b",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "Text 2 provides a relevant explanation for one of the phrases but only addresses the 'forest' and 'trees' metaphor. It fails to mention the 'honey' and 'bees' metaphor, which is also a significant part of the image. It gives ... | The phrases in this image use metaphors to convey motivational messages:
1. **"If you wanna find a honey, you can't be scared of the bees":**
This means that to achieve something valuable or desirable (symbolized by "honey"), you need to be willing to face potential challenges or fears (represented by "bees"). It emphasizes being brave and taking risks to achieve goals.
2. **"If you wanna see the forest, you're gonna have to look past the trees":**
This suggests the importance of seeing the bigger picture rather than focusing too much on small details ("trees"). It encourages perspective and strategic thinking to understand the full scope of a situation.
Both phrases inspire courage and a broader mindset for success. | What do the phrases in the image mean? | [
0.7257745862007141,
0.661888599395752,
0.8253445625305176,
0.8883118629455566
] | 0.007648 | 0.98951 | 0.77533 |
LLaVAR-000000000625 | [
"LLaVA-RLHF-7b-v1.5-224",
"fuyu-8b",
"instructblip-vicuna-13b",
"idefics-9b-instruct"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The response provides an incorrect interpretation of the acronym \"TIP\" and adds an extensive, unfounded explanation which is not helpful in addressing the user's request."
},
"ethical_considerations": {
... | "TIP" in the image stands for the **Technological Institute of the Philippines**, as indicated by the text in the logo. | What does "TIP" stand for in the image? | [
0.8330507278442383,
0.529630720615387,
0.5290228128433228,
0.9224172830581665
] | 0.031345 | 0.919709 | 0.70353 |
llava-detail-680 | [
"llava1.5-7b",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The response is moderately helpful, providing a general description of the scene. It is less detailed than other responses, particularly in describing the individuals and the atmosphere."
},
"ethical_consideration... | This image appears to depict the following features:
1. **Two dogs in motion**: At the center, two dogs seem to be playing or interacting, with one wearing some sort of clothing or outfit.
2. **Indoor setting**: The setting appears to be indoors, possibly a kitchen or dining area, as suggested by the tiled floor and surrounding furniture.
3. **Table with food**: A wooden table is visible to the right, laden with dishes, a green cup, and food like cake or pie on paper plates.
4. **People in the background**: The legs and feet of multiple people (wearing shoes) are visible in the background, suggesting a social gathering.
5. **Blurry action**: The interaction between the dogs is somewhat blurry, indicating motion during the photo capture.
Let me know if you'd like details about a specific part of the image! | Can you describe the main features of this image for me? | [
0.42315706610679626,
0.26924848556518555,
0.30840206146240234,
0.4091099500656128
] | 0.004268 | 0.985638 | 0.352479 |
svit-conversation-862 | [
"VisualGLM-6b",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "Text 4 effectively mentions the aerial trick being performed by the skateboarder, but it is less detailed than Texts 2 and 3. It speaks about the potential popularity of the activity but doesn't provide as much direct observa... | This image shows a skateboarder performing a trick at a skatepark. The person appears to be doing a "stall" or similar maneuver on the edge of a bowl, balancing their skateboard on the coping (the rounded edge of the bowl). The colorful graffiti and skatepark setting add energy to the atmosphere, while others in the background can be seen enjoying the park, either skating, spectating, or relaxing. | What's happening in this image? | [
0.704695463180542,
0.4452285170555115,
0.7215205430984497,
0.6693090200424194
] | 0.012383 | 0.995629 | 0.635188 |
llava-reasoning-776 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The response is helpful as it discusses various issues a beginner might encounter such as balance and control, speed and momentum, timing and coordination, fear and anxiety, physical fitness and endurance, equipment and safet... | A beginner skateboarder could face several challenges in this advanced skatepark setting:
1. **Complex Obstacles**: The ramps, curves, and transitions in this park appear advanced, requiring skills to maintain balance and momentum. Beginners might find it difficult to properly maneuver.
2. **Steep Drops**: These features could be intimidating or lead to loss of control for someone new to skateboarding.
3. **High Skill Requirements**: Performing tricks like ollies or jumps in such technical environments requires confidence and experience that beginners may not have yet.
4. **Safety Risks**: The likelihood of falls, crashes, or injuries increases when skating on such challenging terrain without proper skills or protective gear.
5. **Confidence Levels**: A beginner might feel overwhelmed by the environment, leading to hesitance or a struggle to build confidence.
For safety, beginners are encouraged to practice on simpler, flat surfaces and gradually progress to more advanced parks. | What issue could a beginner skateboarder face in this setting? | [
0.7952690720558167,
0.806187629699707,
0.7558518052101135,
0.8050787448883057
] | 0.00042 | 0.97687 | 0.790597 |
svit-conversation-3480 | [
"LLaVA-RLHF-13b-v1.5-336",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The response is detailed and addresses the instruction effectively by identifying multiple objects and people in the image. It provides additional context to the roles and functions of the subjects within a baseball game, con... | This image contains:
1. Multiple motorcycles parked along the street and some in motion.
2. Riders wearing helmets and riding gear.
3. A road with visible lane markings.
4. A building with signage, including "Herald Sun" and "Bayles General Store," along with Coca-Cola branding.
5. Trees and other vegetation in the background.
6. A phone booth or vending machine near the buildings.
7. A dirt shoulder along the edge of the road. | What objects can you identify in the image? | [
0.43668174743652344,
0.5111267566680908,
0.38597533106803894,
0.2900807559490204
] | 0.006458 | 0.945271 | 0.405966 |
llava-reasoning-4326 | [
"fuyu-8b",
"instructblip-vicuna-13b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The response is moderately helpful because it suggests potential activities that may be taking place, such as meal preparation or multitasking with various ingredients. However, it fails to acknowledge the absence of food or ... | There could be several reasons for having multiple kitchen appliances on the table:
1. **Decluttering and Organizing**: The appliances may have been removed from storage or cabinets to reorganize or clean the kitchen.
2. **Preparation for Use**: These appliances might be set up for a cooking or baking session involving multiple tools.
3. **Testing or Maintenance**: The appliances may need testing, maintenance, or repair to ensure they are functioning properly.
4. **Sorting for Storage or Removal**: The owner could be deciding which appliances to keep or discard, such as for donation or recycling.
5. **Planning a Sale or Giveaway**: The appliances could be organized for sale, such as in a yard sale, or for giving them away.
6. **Limited Space in the Kitchen**: If the kitchen is small or lacks counter space, the table might serve as a temporary spot for these items.
7. **Showcasing or Inventorizing**: The appliances might be displayed for inventory purposes, such as checking what is available or cataloging items for a project.
Understanding the situation involves knowing more about the immediate context! | What might be the reason behind having multiple kitchen appliances on the table? | [
0.719875693321228,
0.8423947095870972,
0.9144551157951355,
0.9745863676071167
] | 0.009002 | 0.996412 | 0.862828 |
LRVInstruction-000000004549 | [
"LLaVA-RLHF-13b-v1.5-336",
"LLaVA-RLHF-7b-v1.5-224",
"fuyu-8b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The text erroneously describes red flowers and bright green foliage, which are not visible in the provided image. While it touches on themes of nature's beauty and preservation, which could be indirectly helpful, it does not ... | The image does not appear to display vibrant red flowers or a plant with bright green foliage. Instead, it features a giraffe partially hidden behind a tree in a grassy environment. The foreground includes green foliage, while the background has a fenced area and stone wall. | Observe the vibrant red flowers blooming on the plant with bright green foliage. They contrast beautifully against the lush green leaves. | [
0.7591120004653931,
0.4531840980052948,
0.5902462005615234,
0.7132588624954224
] | 0.01411 | 0.920968 | 0.62895 |
llava-reasoning-3530 | [
"VisualGLM-6b",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 1 addresses the user's prompt by attempting to describe the challenges a pilot may face. However, it inaccurately assumes the plane is flying low over an open field with obstacles such as buildings, which is not depicted... | In this image, the aircraft is in the process of landing, a critical phase of flight. Several challenges may arise for the pilot during this maneuver:
1. **Crosswinds and Wind Shear**: Crosswinds can cause the aircraft to drift off course, requiring the pilot to continually adjust the inputs. Sudden changes in wind direction or speed (wind shear) can destabilize the descent.
2. **Maintaining Approach Speed**: The pilot must ensure the aircraft is neither too fast nor too slow. Flying too fast could lead to an overshoot, while flying too slow increases the risk of stalling.
3. **Precision in Glide Path**: Staying aligned with the runway at the correct descent angle is crucial. Deviations might result in a hard landing or a missed approach.
4. **Environmental Factors**: Factors such as poor visibility due to weather conditions (fog, rain, or glare) or obstacles (birds, drones) in the landing zone can complicate the process.
5. **Runway Conditions**: Wet, icy, or debris-covered runways could reduce friction, increasing the risk of a runway excursion.
6. **Air Traffic Coordination**: The pilot may need to navigate closely behind other aircraft, which increases the risk of wake turbulence.
7. **Go-Around Decisions**: If the approach becomes unstable, the pilot may have to abort the landing and execute a go-around, which requires timely decision-making and immediate action.
Each of these factors requires the pilot to maintain focus, communicate effectively with air traffic control, and maneuver the aircraft safely. | What challenges might the pilot face during this particular maneuver? | [
0.6743402481079102,
0.553109884262085,
0.6067801117897034,
0.5511143207550049
] | 0.002527 | 0.907551 | 0.596336 |
LLaVAR-000000008821 | [
"VisualGLM-6b",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "Text 4 does not address the user's request. It gives an inaccurate interpretation of the book's content and title, and it fails to mention the authors or translators, thus not contributing positively to the user's comprehensi... | The book mentioned in the image is **"What Is Paleolithic Art? Cave Paintings and the Dawn of Human Creativity"**, written by **Jean Clottes**. It is translated by **Oliver Y. Martin** and **Robert D. Martin**. | What book is mentioned in the image, and who are its authors and translators? | [
0.5658383965492249,
0.1258656084537506,
0.907069206237793,
0.988956868648529
] | 0.115685 | 0.97052 | 0.646933 |
svit-complex_reasoning-8220 | [
"llava1.5-7b",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Text 2 effectively describes the precaution of using an umbrella. It lacks the additional detail of the man's pace, which is a minor omission but still provides a good level of insight."
},
"ethical_considerations... | The man is using a large black umbrella to protect himself from the rain while walking. This is an effective precaution to shield himself from getting wet. | What precautions is the man taking to protect himself from the rain? | [
0.7451575994491577,
0.8278968334197998,
0.7465415000915527,
0.8712514638900757
] | 0.002925 | 0.99155 | 0.797712 |
LRVInstruction-000000009168 | [
"llava1.5-7b",
"fuyu-8b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Text 3 directly answers the user's question by confirming that there are no water slides or diving boards, but it does not provide additional context or description that might enhance the user's understanding, as Text 1 does.... | The pool shown in this image does not have visible water slides or diving boards. It appears to be a standard recreational pool situated in a pleasant setting with nearby seating and palm trees. | Does the swimming pool have water slides and diving boards? | [
0.7194784283638,
0.7347371578216553,
0.66997230052948,
0.9196746945381165
] | 0.008969 | 0.963865 | 0.760966 |
LLaVAMed-000000000225 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The response provides a detailed explanation of what nodules may look like on a CT scan and the potential health implications they may indicate. It also emphasizes the importance of a medical professional's evaluation for dia... | In this CT scan image, a nodule would appear as a circular or oval-shaped area that is more defined than the surrounding tissue. Its appearance might vary depending on its size, location, and density. From this scan:
- **Liver nodules (if visible):** These might appear as round or irregular areas differing in density compared to the normal liver tissue.
- **Other nodules:** Areas of altered density in other soft tissues or organs could also represent nodules.
For accurate interpretation and clinical relevance, a radiologist would need to analyze the scan in more detail, considering the context, patient history, and the contrast phases. If this scan shows abnormalities, they might require further analysis to determine their nature. | What do the nodules look like on the CT scan? | [
0.8306195735931396,
0.59275221824646,
0.8088141679763794,
0.8662539720535278
] | 0.011444 | 0.994141 | 0.77461 |
LLaVAR-000000004887 | [
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"idefics-9b-instruct",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response simply names the book and author, providing minimal information and no description of the visual elements, which is not helpful in the context of the task."
},
"ethical_considerations": {
"Ra... | The painting in the image depicts an abstract view of a harbor. It uses shades of blue and white to convey the ocean and sky, with sweeping lines and strokes that illustrate movement and energy. The artwork features a prominent bridge, possibly resembling a famous structure like a Sydney Harbour Bridge, as well as curved lines suggesting roads or pathways surrounding the water. The dynamic, slightly impressionistic style creates a lively and textured representation of the harbor's scenery. | Can you describe the painting in the image? | [
0.49964576959609985,
0.7380480766296387,
0.804917573928833,
0.6148850321769714
] | 0.013691 | 0.954054 | 0.664374 |
LRVInstruction-000000007608 | [
"LLaVA-RLHF-7b-v1.5-224",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "This text is helpful as it not only confirms the location of the bench but also provides context about its purpose and relationship to the skateboarding area."
},
"ethical_considerations": {
"Rating": "5",... | The image shows a person skateboarding on a ledge within a skating area, but there isn't a visible bench next to the skating area in this scene. Let me know if there's something specific you'd like to find! | Find the bench next to the skating area. | [
0.6793552041053772,
0.4323592782020569,
0.7660691738128662,
0.6876160502433777
] | 0.015704 | 0.995211 | 0.64135 |
svit-complex_reasoning-2009 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-7b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "The response accurately identifies brown as the most prominent color in the room and provides a detailed description of how the color contributes to the room's design and atmosphere. It is helpful as it addresses the user's q... | The most prominent color in this room is shades of **brown**, which is visible in the wooden furniture, flooring, and detailing, creating a warm, earthy tone throughout the space. | What is the most prominent color in this room? | [
0.9097152352333069,
0.7841062545776367,
0.5427114963531494,
0.9278921484947205
] | 0.023632 | 0.997556 | 0.791106 |
svit-complex_reasoning-8546 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-13b",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "This response is moderately helpful. It correctly identifies the most common color and provides context about the environment but makes additional assumptions about the buildings and the area's busyness which were not request... | The most common color for the signs in the image appears to be **red**, as it is prominently featured in many of the illuminated advertisements throughout the scene. | What is the most common color for the signs as observed in the image? | [
0.6889137029647827,
0.7730117440223694,
0.8245647549629211,
0.8577338457107544
] | 0.004057 | 0.995538 | 0.786056 |
llava-reasoning-4109 | [
"VisualGLM-6b",
"instructblip-vicuna-7b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 2 only tangentially addresses the prompt by focusing on the potential implications of alcohol consumption, which is not directly evident or central to the image. It offers some relevant information on health and social i... | There are several potential issues in the environment shown in this image:
1. **Clutter and Safety Hazards**: The table in the image has various items (glasses, bottles, and a remote), creating clutter that could be knocked over, potentially causing spills, breakage, or injury.
2. **Distractions**: The scattered items and general untidiness of the area may make it harder for the individual to focus on the game or other activities.
3. **Comfort and Posture**: While lounging can be relaxing for a short time, prolonged sitting or reclining in this position could lead to discomfort or strain on the back and neck over time.
4. **Food or Drink Spills**: Having beverages near electronics like the television and gaming console poses a risk of accidental spills, which could damage the equipment.
5. **Fire or Electrical Hazard**: If the items on the table obstruct airflow to electrical devices (if chargers or wires are underneath), it could increase the risk of overheating.
6. **Eye Strain**: Sitting too close to the screen for an extended period could strain the eyes, especially in low lighting.
7. **Potential Poor Organization**: The visible mess may make it difficult to find items or maintain the space, leading to further disorganization and mental stress.
Improving this setup—by organizing the space, clearing clutter, and ensuring ergonomic seating—could make gaming more enjoyable and safe! | What potential issues are there in the environment surrounding the person playing video games? | [
0.47978031635284424,
0.7027674317359924,
0.7180763483047485,
0.7839899063110352
] | 0.013139 | 0.963327 | 0.671154 |
LLaVAMed-000000001009 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-13b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The text inaccurately describes the image as showing a double aortic arch, which is not visible or indicated in any way in the provided image. This misinterpretation does not help the user in understanding the image."
}... | The image shows a cross-sectional MRI of the chest with labels indicating the organs and structures. The deformity indicated in the image is **Pectus Excavatum**, a condition where the sternum (xiphisternum) is sunken inward towards the spine, leading to a concave or "caved-in" appearance of the chest wall.
### Key features seen in the image:
- The **xiphisternum** is markedly depressed inward, into the thoracic cavity.
- This displacement can compress the heart and other mediastinal structures, which is evident as the heart appears shifted or compressed.
Pectus Excavatum is the most common chest wall deformity and may result in functional impairments or cosmetic concerns depending on the severity. | What is the deformity shown in the image? | [
0.4793491065502167,
0.3442143499851227,
0.6339279413223267,
0.6384237408638
] | 0.014873 | 0.91593 | 0.523979 |
PMC-VQA-301 | [
"VisualGLM-6b",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 2 seems to have misunderstood the content of the image, assuming it's related to blood vessels and the health of mice, which does not align with the visual evidence or the implied context of the experiment."
},
... | The endpoint of the experiment described in part A was the measurement of tumor volume. The graph shows tumor volumes (in mm²) for three samples (no. 1, no. 2, and no. 3) in the control group (shCtl) and the experimental group with PDCD4 knockdown (shPDCD4). The results indicate increased tumor volume in the shPDCD4 group compared to the shCtl group, suggesting that PDCD4 knockdown promotes tumor growth. | What was the endpoint of the experiment described in part A? | [
0.3405572772026062,
0.050257354974746704,
0.34237343072891235,
0.6572144627571106
] | 0.046088 | 0.941899 | 0.347601 |
svit-complex_reasoning-3968 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The response begins with incorrect information about the color of the shoes, which is not yellow. It provides additional context, but since the initial information is incorrect and the additional details were not requested, i... | The skateboarder's shoes are blue and white. | What is the color of the skateboarder's shoes? | [
0.5517879724502563,
0.9070843458175659,
0.17450985312461853,
0.8818978071212769
] | 0.088453 | 0.983656 | 0.62882 |
svit-conversation-3282 | [
"LLaVA-RLHF-13b-v1.5-336",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The response provides an extended narrative that isn't entirely supported by the image content. It makes assumptions about the environment and the background activity that cannot be confirmed by the image, which detracts from... | The image shows four individuals seated at wooden booths in a casual restaurant setting. Two people, who appear to be men, are sharing a table on the right. The man in the red tie is eating, while the other, seen from the back, is engaging with his food as well. On the left side of the image, two women are seated at a separate booth, facing each other and conversing. One woman is wearing a turquoise outfit, and the other is wearing a patterned top. Both groups appear to be enjoying a meal in a relaxed and friendly atmosphere. The restaurant interior features exposed brick walls and large windows. | Can you describe the people present in the image? | [
0.7265454530715942,
0.8349496126174927,
0.7569823861122131,
0.8523178100585938
] | 0.002748 | 0.966257 | 0.792699 |
llava-conversation-1420 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The response correctly identifies the color of the bus as blue and white, providing additional context regarding the black stripe that might suggest its function, which aligns with the user's request for color identification ... | The bus in the image is orange. | What color is the bus in the image? | [
0.6338691711425781,
0.6436215043067932,
0.6436215043067932,
0.6436215043067932
] | 0.000018 | 0.991966 | 0.641183 |
m3it-vist-2 | [
"VisualGLM-6b",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "This text provides a descriptive narrative that aligns well with the image, giving context to the vehicles and suggesting their potential uses. However, it identifies one vehicle as a police car, which is not confirmed by the... | In a futuristic desert town, self-driving rally cars prepared for an epic race to test cutting-edge AI technology. Among them, a blue Volkswagen and two red Caterpillar jeeps stood out, covered in sensors and powerful hardware. The rivals, Team Blue and Team Red, engaged in a fierce battle across rocky landscapes and dusty flatlands. After encountering challenging hazards, Team Blue’s advanced navigation AI triumphed over the treacherous terrain. Celebrating at the finish line, the winning team marveled at how innovation had driven both competition and collaboration. | Given five images, write a coherent five-sentence story about the images. | [
0.39982470870018005,
0.20087800920009613,
0.3966479003429413,
0.3368200659751892
] | 0.006497 | 0.998859 | 0.333543 |
svit-conversation-4213 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-13b",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The text provides relevant information about the boat and its surroundings, suggesting it might be an active port or marina. However, the mention of several cars nearby is incorrect, as there are no cars visible in the image,... | This image shows an urban intersection with several distinct elements:
1. **Railroad Crossing:** The foreground includes a railroad crossing with raised tracks, safety barriers, and warning signals (lights and crossbucks).
2. **Traffic Signals:** There are traffic lights controlling vehicle flow at the intersection.
3. **Bus:** A white city bus is stopped along the street, with visible advertisements.
4. **Businesses:** There are buildings with signs, such as a "Noodles & Co." restaurant in a brick building on the left and other businesses visible on the right.
5. **Street Scene:** The street has vehicles, pedestrians in the distance, and other small details like crossing signals and trees.
6. **Weather:** The sky is overcast, with gray clouds suggesting cloudy or dreary weather.
It appears to be a busy urban or downtown setting. | What can you see in the image? | [
0.49163416028022766,
0.5198448896408081,
0.30239051580429077,
0.3941133916378021
] | 0.007352 | 0.9635 | 0.426996 |
llava-conversation-2265 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-13b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "This response is highly helpful as it correctly identifies the fruits in the blender and elaborates on their health benefits and potential uses in a smoothie, providing comprehensive and relevant information that enhances und... | The blender contains the following fruits:
- Bananas (sliced)
- Oranges (peeled)
- Strawberries
- Apple (small pieces) | What types of fruits are in the blender? | [
0.6589481830596924,
0.9126497507095337,
0.8739907741546631,
0.898584246635437
] | 0.010646 | 0.964121 | 0.836043 |
svit-complex_reasoning-8215 | [
"LLaVA-RLHF-13b-v1.5-336",
"instructblip-vicuna-13b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "Text 1 provides a detailed analysis of various elements that might indicate the setting is a museum or a place of historical significance. However, it inaccurately mentions a \"glass case or barrier\" and a \"small sign or la... | Several elements of the image suggest that it may be set in a museum or a place of historical significance:
1. **Display Setup**: The vase is placed on a clear stand within a controlled display, typical of museum exhibitions. The surrounding objects also seem to be similarly arranged for viewing.
2. **Artifacts**: The vase appears to be an ancient or historical artifact, such as a Greek or Roman ceramic, suggesting that it is being preserved and showcased for educational or cultural purposes.
3. **Neutral Background**: The plain, neutral-colored wall behind the objects highlights the artifacts and resembles the setup commonly used in museum displays to avoid distractions.
4. **Lighting**: The lighting appears to be uniform and strategic, emphasizing the details of the artifacts, which is standard in museum contexts.
Altogether, these features strongly suggest the image is set in a museum or a curated location focused on historical preservation. | What elements of the image suggest it may be set in a museum or a place of historical significance? | [
0.8794611692428589,
0.807577908039093,
0.8241574764251709,
0.8273171782493591
] | 0.000726 | 0.971048 | 0.834628 |
llava-conversation-3118 | [
"LLaVA-RLHF-13b-v1.5-336",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "This response provides a detailed explanation that not only identifies the motorcycles but also gives context about motorcycle events and their potential implications for the local community. It gives more than what is asked ... | The vehicles parked along the city street in this image are motorcycles. Many of them have shiny finishes, flags, and accessories, and appear to be part of a gathering or event. | What type of vehicles are parked along the city street? | [
0.6662631034851074,
0.8750969171524048,
0.7991966009140015,
0.8870969414710999
] | 0.00773 | 0.994116 | 0.806913 |
PMC-VQA-48 | [
"llava1.5-7b",
"llava1.5-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The response mentions the optic nerve head, which is not visible in the image, leading to potential confusion. While it correctly identifies the image as a cross-section of the retina, the information is not fully accurate or... | The bottom left image presents an **en face OCT angiography (OCTA) scan** of a retinal or choroidal vascular layer. In this image:
- The **yellow dashed circle** highlights an area of interest, likely corresponding to abnormal vasculature or a lesion.
- The vascular structures are visualized in detail, showing areas of blood flow in enhanced contrast, which could indicate pathology such as **choroidal neovascularization (CNV)**.
- Compared to the ICGA series above, the OCTA provides a detailed, non-invasive view of the vasculature without needing dye and highlights the blood vessels’ morphology.
The en face scan specifically shows a "top-down" perspective of blood flow, offering a planar view of the selected layer's microvasculature that corresponds with the circular area of interest. | What is shown in the en face OCTA scan at the bottom left of the image? | [
0.6647740602493286,
0.6261856555938721,
0.17609046399593353,
0.8665883541107178
] | 0.063637 | 0.970183 | 0.58341 |
svit-detail_description-462 | [
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b",
"instructblip-vicuna-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 3 contains relevant information about the scene, such as the presence of a tennis player, spectators, and the setting. However, it inaccurately describes the player's attire and incorrectly refers to the player as a man.... | The image shows a close-up of a person holding scissors and a roll of tape. Here are the details:
1. **Person**: A person is partially visible in the background with their face somewhat obscured by the objects they are holding. Their hair is pulled back, and they appear to be indoors.
2. **Scissors**: The person is holding a shiny pair of scissors with a black handle in one hand. The blades are metallic and reflective.
3. **Roll of Tape**: A roll of clear tape with a red tint on its edges is held in the other hand. The inner cardboard of the tape roll is visible.
4. **Background**: The background is a red pattern with white triangular designs that resemble stylized Christmas trees.
5. **Lighting**: The lighting emphasizes the person’s hand holding the scissors and tape. It creates a warm glow, reflecting off the metal blades of the scissors.
6. **Focus**: The focus of the image seems to be on the hands and the objects being held rather than on the person's face, which appears slightly blurred.
This image likely suggests a crafting or preparation activity, possibly related to wrapping or holiday decor due to the Christmas-themed background. | Disclose every detail you see in the image. | [
0.3605504333972931,
0.24249958992004395,
0.3584752678871155,
0.14470338821411133
] | 0.008078 | 0.925242 | 0.276557 |
m3it-textcap-11 | [
"llava1.5-7b",
"llava1.5-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The response identifies the main subject (a book by Dylan Thomas) but fails to mention that it's an album cover, which is an important context that contributes to understanding the image."
},
"ethical_consideratio... | The image displays a stack of copies of the book *Things Fall Apart* by Chinua Achebe. | Write a one-sentence description of the image, which would require reading the text in the image. | [
0.28584709763526917,
0.5004966259002686,
0.35804688930511475,
0.5641701221466064
] | 0.012224 | 0.992409 | 0.42714 |
svit-conversation-413 | [
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response is not helpful as it lists a wide array of objects that are not present in the image. It provides an overwhelming and inaccurate list that does not contribute to the user's understanding of this specific image."... | This image includes:
- A red flower with a long stem.
- A red vase holding the flower.
- A lamp with a clear, faceted glass or crystal base and a beige lampshade. | What objects are present in the image? | [
0.34916266798973083,
0.18697747588157654,
0.20087431371212006,
0.2404813915491104
] | 0.004046 | 0.872433 | 0.244374 |
LLaVAMed-000000003532 | [
"llava1.5-7b",
"llava1.5-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The response is moderately helpful as it provides a general description of the image and identifies that it is a representation of the human ribcage and lungs. However, it lacks the specificity and detail that could enhance t... | This is a coronal cross-sectional medical image, most likely a CT scan of the chest, showing the lungs and part of the surrounding thoracic structures. Here's what can be observed:
1. **Lung Fields**:
- The darker areas represent air-filled lung tissue (low density).
- The uneven texture of the lungs suggests that this image might show abnormalities such as scarring, infiltrates, or other conditions affecting lung parenchyma.
2. **Trachea and Bronchi**:
- The central dark structure branching into the lungs represents the trachea and bronchi.
3. **Surrounding Anatomy**:
- The gray, denser structures on the periphery represent soft tissue and bones (such as ribs and vertebrae).
4. **Possible Pathology**:
- The image hints at potential damage or structural abnormality in the lungs (e.g., severe inflammation or fibrosis), though a radiologist would need to analyze it further for a definitive diagnosis.
If this question is for medical purposes, it is important to consult a healthcare professional or radiologist for an accurate interpretation and advice. | What can you tell me about the image? | [
0.6905364990234375,
0.7144304513931274,
0.7220131158828735,
0.8981380462646484
] | 0.006843 | 0.993125 | 0.75628 |
m3it-st-vqa-12 | [
"VisualGLM-6b",
"fuyu-8b",
"instructblip-vicuna-7b",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The response is not helpful as it fails to address the user's specific question about delivery costs. Instead, it provides irrelevant information about the potential cost of a slice or meal and suggests contacting the establi... | The cost for delivery of Espressa Pizza is not specified in the visible signage in this image. It would likely require directly contacting the shop via the phone number provided on their signage. | Given the image, answer the question related to the visual content. What is the cost for delivery of Espressa Pizza? | [
0.887627124786377,
0.0536009706556797,
0.0911666676402092,
0.23360633850097656
] | 0.113237 | 0.685121 | 0.3165 |
svit-detail_description-743 | [
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "Text 2 is moderately helpful, providing a description of the man's attire and the items on the table. However, it includes inaccuracies, such as a \"white tablecloth\" and a \"bowl of fruit,\" which are not visible in the ima... | This image showcases a scene with two people walking in the foreground while a group of horses and riders are visible in the background, likely at an outdoor event or gathering. Here's a detailed breakdown of the nuances:
1. **People in the Foreground**:
- The two individuals are walking side by side, suggesting they are together and possibly engaged in conversation or a ceremonial activity.
- The woman holds a bouquet of flowers, hinting at a celebratory or formal occasion.
- Both individuals are dressed casually yet neatly, indicating a rural or outdoor context that balances formality and practicality.
2. **Horses in the Background**:
- A line of horses, each with a rider, is visible behind the pair. The horses are equipped with saddles and decorated tack, suggesting they are prepared for a performance, ride, ceremony, or parade.
- The riders' attire and the decorative nature of the tack might imply this is part of a special event rather than a regular horseback ride.
3. **Environment**:
- The scene takes place on a dirt surface, suggesting the location is an arena, ranch, or open rural space.
- The atmosphere feels grounded and communal, appropriate for outdoor equestrian events or local gatherings.
4. **Nuances of Interaction**:
- The walking pair seems calm and engaged, with their demeanor conveying purpose, respect, or involvement in an organized event.
- The bouquet hints at a gesture of appreciation, honor, or recognition.
5. **Cultural and Contextual Clues**:
- The presence of horses and traditional elements implies a cultural or regional activity, possibly involving ranching, a rodeo, or a local festival.
- The mix of modern and traditional elements (such as clothing styles and horse gear) emphasizes the balance of tradition and contemporary culture in this setting.
The composition and body language convey a harmonious moment, creating a snapshot of shared community or ceremony. | Explain all the nuances you observe in the image. | [
0.3154274821281433,
0.0892581194639206,
0.20226196944713593,
0.5456300377845764
] | 0.028494 | 0.926158 | 0.288144 |
svit-complex_reasoning-9983 | [
"VisualGLM-6b",
"idefics-9b-instruct",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The response correctly identifies the player as a batter but then makes an unfounded assumption about multiple teams playing throughout the day, which is not something that can be inferred from the image. This additional info... | The player wearing the black helmet in the foreground is holding a bat and standing near the dugout, likely in the on-deck circle. This suggests that they are a batter preparing to hit and likely play as an offensive position, such as an outfielder, infielder, or designated hitter. | What position does the player wearing the black helmet likely play? | [
0.8491133451461792,
0.8180211782455444,
0.8883415460586548,
0.8950363397598267
] | 0.000971 | 0.991026 | 0.862628 |
LRVInstruction-000000001052 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"fuyu-8b",
"instructblip-vicuna-13b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "The response incorrectly assumes details about the clock that are not visible in the image, such as the time it displays. It also incorrectly assumes the time of day and the girl's actions, which were not part of the user's q... | There is no clock visible in this image, so I cannot determine whether the time is correct. | Does the clock on the wall show the correct time? | [
0.6364198327064514,
0.6860289573669434,
0.1125900149345398,
0.730383038520813
] | 0.062385 | 0.912581 | 0.541355 |
LRVInstruction-000000005618 | [
"fuyu-8b",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response is not helpful because it inaccurately affirms the presence of a dog running in the snow, which does not correspond to the image content."
},
"ethical_considerations": {
"Rating": "4",
... | There doesn’t appear to be a dog running in the snow in this image; the focus is on a skier on a snowy mountainside. Let me know what else you'd like help with! | Can you spot the dog running in the snow? | [
0.6573962569236755,
0.6424045562744141,
0.016007931903004646,
0.7126184105873108
] | 0.081076 | 0.835721 | 0.507107 |
svit-complex_reasoning-1635 | [
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response is largely unhelpful; it provides a confused analysis with incorrect observations (e.g., the pizza is on a paper plate, not an aluminum pan), irrelevant information (e.g., speculation about cheese types and rest... | This pizza slice appears to have the following toppings:
- Black olives
- Spinach or basil (green leaves)
- Red sauce or hot sauce (drizzled)
- Shredded cheese
- Red pepper flakes (sprinkled)
Let me know if you need anything else! | What are the toppings on the pizza slice? | [
0.7228549122810364,
0.7198059558868408,
0.6814486384391785,
0.7834905385971069
] | 0.001334 | 0.961364 | 0.7269 |
LRVInstruction-000000006635 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"MMICL-Vicuna-13B",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The text is moderately helpful as it correctly identifies the attire of the person on the motorcycle and attempts to provide context regarding the appropriateness and psychological impact of the attire in the setting. However... | The man on the motorcycle in the image is indeed wearing a white and red outfit, including what appears to be a motocross-style jacket. | The man on the motorcycle is wearing a white and red coat. | [
0.6757633686065674,
0.7983922362327576,
0.42361319065093994,
0.8350856304168701
] | 0.025945 | 0.987927 | 0.683214 |
svit-complex_reasoning-2259 | [
"LLaVA-RLHF-13b-v1.5-336",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "Text 1 provides a list of possible reasons why a plane might have folded wings. However, most of the reasons listed are not applicable to the type of aircraft shown in the image, such as the suggestion that folding the wings ... | This plane has its wings folded to save space, a feature commonly used on aircraft designed for use on aircraft carriers. The folding wing capability allows more planes to be stored in the confined space of a carrier's hangar deck. It also aids in maneuvering the aircraft on the carrier’s deck when preparing for or after operations. This feature is typical of many naval aircraft, like the Grumman TBF Avenger (seen in the image), to enhance the logistical efficiency of carrier-based aviation. | Why might the plane have its wings folded? | [
0.7000601291656494,
0.6376744508743286,
0.7638051509857178,
0.7933307886123657
] | 0.003604 | 0.978141 | 0.723718 |
svit-complex_reasoning-3682 | [
"instructblip-vicuna-7b",
"idefics-9b-instruct",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Similar to text 1, this text effectively describes the action in the image. It is helpful and addresses the instruction correctly."
},
"ethical_considerations": {
"Rating": "5",
"Rationale": "This ... | The woman in the image appears to be comforting or checking the dog, possibly inspecting or soothing it. She has one arm around the dog and seems to be tending to it in a caring manner. | What is the woman in the image doing with the dog? | [
0.6995357871055603,
0.6555227041244507,
0.6995357871055603,
0.9587322473526001
] | 0.014386 | 0.996967 | 0.753332 |
svit-complex_reasoning-524 | [
"LLaVA-RLHF-13b-v1.5-336",
"fuyu-8b",
"idefics-9b-instruct",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The response goes into detail about what the act of one girl talking on the phone while both smile might suggest about their experience and the role of cell phones in social interaction. It also touches on the emotional and s... | One of the girls is smiling while holding a phone to her ear. | What is one of the girls doing while they are both smiling? | [
0.6505066752433777,
0.3983219265937805,
0.7350280284881592,
0.8748923540115356
] | 0.030071 | 0.973963 | 0.664687 |
LLaVAR-000000001922 | [
"llava1.5-7b",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The text is helpful and directly addresses the user prompt by interpreting the message and relating it to the overall well-being. It adds the observation of the orange background contributing to the message's uplifting nature... | The message "Be More Well" conveys a call to focus on wellness and self-improvement. It encourages individuals to enhance their mental, physical, and emotional health. The minimalist presentation of the text against a bright background suggests simplicity and positivity, aligning with themes of mindfulness and overall well-being. | What message does the image convey with the text "Be More Well"? | [
0.8415235877037048,
0.8757688999176025,
0.8473616242408752,
0.902417778968811
] | 0.000591 | 0.998753 | 0.866768 |
LLaVAMed-000000003986 | [
"LLaVA-RLHF-13b-v1.5-336",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "This text is highly helpful, providing a detailed explanation of what the resected part is, the context of a surgical procedure, and potential reasons for such a procedure. It adds value by explaining the implications of the ... | The resected part in the image is a **meningioma**, a type of brain tumor originating in the meninges, which are the protective membranes covering the brain and spinal cord. The images on the left (J and K) show the excised tumor tissue, and the image on the right is an axial CT scan showing the location of the tumor before excision. Meningiomas are generally slow-growing and are often surgically removed if symptomatic. | What is the resected part in the image? | [
0.6779040098190308,
0.5763789415359497,
0.7573215961456299,
0.6967341899871826
] | 0.004242 | 0.981562 | 0.677085 |
LLaVAR-000000000661 | [
"fuyu-8b",
"instructblip-vicuna-13b",
"idefics-9b-instruct",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "The response correctly provides the quote and the author, along with the lifespan years, which corresponds exactly to the user's instruction."
},
"ethical_considerations": {
"Rating": "5",
"Rationa... | The quote on the image is:
"The goal isn’t to live forever, but to create something that will."
It is attributed to Selena Quintanilla Pérez (1971-1995). | What is the quote on the image and who said it? | [
0.8197183609008789,
0.40957778692245483,
0.9079470038414001,
0.9985718131065369
] | 0.050718 | 0.99661 | 0.783954 |
LRVInstruction-000000008723 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"fuyu-8b",
"instructblip-vicuna-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "This text attempts to address the presence of a red barn, which aligns with the instruction. However, it introduces speculative information not present in the image, such as a fenced-in area and the barn's purpose for safety,... | There does not appear to be a red barn visible in the background of this image. Only the sheep, green grass, and trees can be seen. Let me know if you'd like help enhancing or editing the image! | Watch for a red barn in the distance behind the tree. | [
0.6618597507476807,
0.5588113069534302,
0.39906975626945496,
0.21682585775852203
] | 0.028339 | 0.883556 | 0.459142 |
svit-complex_reasoning-315 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"fuyu-8b",
"idefics-9b-instruct"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response provides a confused identification (\"sandstorm or beached bird\"), which is not only incorrect but also does not offer useful information about the bird or its environment."
},
"ethical_consideratio... | The bird in the image appears to be a **sandpiper**, a type of shorebird commonly found along coastlines. More specifically, it may be a **Sanderling** or another small species of sandpiper, as it is wading on the shore, which is characteristic behavior of these birds. | What type of bird is featured in the image? | [
0.8155108690261841,
0.6785666942596436,
0.6388887166976929,
0.6854867935180664
] | 0.004416 | 0.947612 | 0.704613 |
llava-conversation-734 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"fuyu-8b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "2",
"Rationale": "This response identifies the bird as a seagull but includes a confusing and incorrect statement about the bird's positioning, suggesting it might be \"sitting in a window or at an angle,\" which is not supported by the image ... | The bird perched on the car's rearview mirror is a type of gull, likely a seagull. This species is common along coastal areas and is known for its light-colored feathers, long beak, and webbed feet. | What type of bird is on the car's rearview mirror? | [
0.8567066192626953,
0.8158208727836609,
0.41262924671173096,
0.8064396977424622
] | 0.032446 | 0.945564 | 0.722899 |
LLaVAR-000000000833 | [
"VisualGLM-6b",
"LLaVA-RLHF-13b-v1.5-336",
"LLaVA-RLHF-7b-v1.5-224",
"MMICL-Vicuna-13B"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "Text 3 fails to address the theme of the book entirely, instead incorrectly describing the image without providing any information related to the book's content or theme."
},
"ethical_considerations": {
"R... | The theme of "The Spectre of Defeat in Post-War British and US Literature: Experience, Memory, and Post-Memory," edited by David Owen and Cristina Pividori, revolves around the exploration of the concept of defeat and its lingering effects in British and American literature following the wars. It delves into themes of memory, post-memory, trauma, and cultural representation, analyzing how the specter of defeat influences narratives, societal consciousness, and individual identities in literary works. The book examines how writers process and articulate the experiences of loss, failure, and their aftermath in the context of war and its broader implications. | What is the theme of the book "The Spectre of Defeat in Post War British and US Literature"? | [
0.25325441360473633,
0.9116435647010803,
0.9322859644889832,
0.6340785026550293
] | 0.075374 | 0.983003 | 0.682816 |
svit-complex_reasoning-6625 | [
"LLaVA-RLHF-13b-v1.5-336",
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-7b",
"idefics-9b-instruct"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "This response is moderately helpful, providing insights into how sunny weather affects a motorcyclist's experience. However, it includes speculative information about the motorcyclist's clothing choices and well-being, which ... | The weather in the image appears to be partly cloudy with a mix of blue sky and white clouds. Visibility is good, and there are no signs of precipitation or strong winds on the water. It seems relatively calm and pleasant. | What is the weather like in the image? | [
0.4023326337337494,
0.536986231803894,
0.6195738315582275,
0.47820407152175903
] | 0.006334 | 0.992766 | 0.509274 |
LRVInstruction-000000009957 | [
"fuyu-8b",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "This response directly addresses the user's question by confirming the absence of a dinosaur and providing a description of what is actually in the image, which is relevant and helpful for the user's understanding."
},
... | No, there is no gigantic purple dinosaur in the image. It shows a cityscape with a river, bridge, trees, buildings, and a yellow tram, but no dinosaur. | Is there a gigantic purple dinosaur standing next to the trees? | [
0.7594342231750488,
0.038804613053798676,
0.1978626251220703,
0.9738322496414185
] | 0.148896 | 0.944133 | 0.492483 |
svit-complex_reasoning-8493 | [
"VisualGLM-6b",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The response provides incorrect color information about the letters on the bus, stating they are green with a blue background, which is not evident in the image provided."
},
"ethical_considerations": {
"R... | The letters on the bus are primarily orange, yellow, and white. | What is the color of the letters on the bus? | [
0.7164257764816284,
0.5993678569793701,
0.7908027172088623,
0.7908027172088623
] | 0.006129 | 0.998511 | 0.72435 |
LRVInstruction-000000003459 | [
"llava1.5-7b",
"fuyu-8b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The text incorrectly references a chandelier and a bed in the setting, which are not present in the image. It fails to correct the prompt's inaccuracies, providing no valuable insight into the actual content of the image, ren... | It seems the image depicts a worn-out or aged setting, such as a cell or basic restroom, rather than an elegant chandelier. The space evokes a sense of aged simplicity, with muted lighting casting shadows on the worn walls and fittings. The ambiance leans more toward gritty and historical than cozy or luxurious. If you're envisioning a chandelier's elegance, this setting provides an interesting contrast! | Notice how the elegant chandelier hanging from the ceiling illuminates the room with a warm and cozy light. | [
0.751428484916687,
0.7920811176300049,
0.712971031665802,
0.8161693811416626
] | 0.001551 | 0.960791 | 0.768163 |
LRVInstruction-000000005677 | [
"LLaVA-RLHF-7b-v1.5-224",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "This response is helpful as it not only confirms that the propellers are spinning but also provides an explanation of what this indicates for the helicopter (in flight, generating lift and thrust). It adds context about the r... | Yes, in the image, the propellers on top of the helicopter appear to be spinning. The tips of the blades show a slightly blurred motion, which is common for a helicopter in flight. However, they are not distinctly red; they appear mostly dark with the tips possibly being colored. The main rotor blades are clearly visible. | Can you observe the red propellers spinning on top of the helicopter? | [
0.8846798539161682,
0.7403676509857178,
0.8936254382133484,
0.8835855722427368
] | 0.004063 | 0.982942 | 0.850565 |
svit-complex_reasoning-8745 | [
"VisualGLM-6b",
"instructblip-vicuna-7b",
"idefics-9b-instruct",
"Qwen-VL-Chat"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "The response is helpful and provides a good explanation of what the boy might be doing, mentioning playtime and learning concepts. It loses one point for slightly less clarity in the statement about the boy possibly preparing... | Based on the boy's positioning and the play kitchen setup he is interacting with, it appears he is engaged in imaginative play, simulating cooking or preparing food. He might be pretending to take something out of the refrigerator, bake, or organize items, which is a common activity for children in such an environment. | Based on the way the boy is positioned and what he's interacting with, what activity could he be engaged in? | [
0.8313681483268738,
0.8054299354553223,
0.8124412298202515,
0.8196200132369995
] | 0.000092 | 0.997738 | 0.817215 |
LRVInstruction-000000007051 | [
"VisualGLM-6b",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "Text 2 incorrectly states that the car is floating and approaching a traffic light, which is not present in the image. It also misinterprets the scene by suggesting that the car is for display purposes, which is not supported... | No, the older car near the stop sign is not floating above the ground. It may appear as though it is due to the angle of the photograph, lighting, or shadows, but the car is resting on its tires and is grounded on the road. | Is the older car at the stop sign floating above the ground? | [
0.7607343196868896,
0.663371205329895,
0.9050893783569336,
0.9308549165725708
] | 0.011869 | 0.978233 | 0.815012 |
LRVInstruction-000000009248 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-13b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "3",
"Rationale": "The text provides a detailed description of the urban setting but only moderately addresses the instruction about the treasure chest, noting its absence. It gives additional context about the image, which may be useful to the... | This seems to be a street scene in a Chinatown or similarly styled urban area, bustling with businesses and shops. If you are referring to a hidden object, it is likely a metaphorical or playful statement. The area under the orange and blue sign could be checked, but as with any such claim, it's wise to approach urban treasure hunting with caution and respect for private property and local regulations. | There's a hidden treasure chest under the orange and blue sign. | [
0.6824246644973755,
0.37616240978240967,
0.44166386127471924,
0.4821016192436218
] | 0.013065 | 0.981855 | 0.495588 |
svit-conversation-3925 | [
"fuyu-8b",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The text is not helpful as it inaccurately describes an object that is not visible in the image. There is no advertisement visible on the bus."
},
"ethical_considerations": {
"Rating": "5",
"Ration... | The bus in the image is labeled as the "61 Shuttle," and its electronic display indicates it's route "27," traveling to "Reading Friar St" and "Caversh." The bus is parked at a bus stop on what appears to be a street in a small town or urban locale. | What's on the bus? | [
0.5991716980934143,
0.11735941469669342,
0.5661730766296387,
0.8174116611480713
] | 0.064718 | 0.930436 | 0.525029 |
llava-conversation-3001 | [
"llava1.5-7b",
"llava1.5-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "4",
"Rationale": "Similar to Text 1, this response is helpful and correctly identifies the color of the bowl but lacks the detail of the plate's rim color, which is a significant part of the visual."
},
"ethical_considerations": {
... | The bowl is dark blue, and the plate is green. | What color is the bowl and plate in the image? | [
0.896648645401001,
0.896648645401001,
0.9108930826187134,
0.8493651151657104
] | 0.000541 | 0.992517 | 0.888389 |
svit-conversation-2988 | [
"VisualGLM-6b",
"LLaVA-RLHF-7b-v1.5-224",
"MMICL-Vicuna-13B",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This text inaccurately describes the image, mistaking it for a public park or community center and referring to an electric power pole and building or school that are not clearly depicted. The translation of \"铁丝网\" (wire mes... | The image shows an industrial or secured facility surrounded by a chain-link fence topped with barbed wire, indicating restricted access. Several signs are present, including warnings about trespassing prosecution and security measures. A vintage-style sign reads "N.Y.C.T.A. Structures," which suggests it is related to a New York City Transit Authority property. The scene portrays a heavily protected area, likely for transportation or government operations. | What does the image depict in general? | [
0.6726376414299011,
0.6807078123092651,
0.6269035339355469,
0.8245823979377747
] | 0.005495 | 0.991893 | 0.701208 |
llava-reasoning-4824 | [
"LLaVA-RLHF-7b-v1.5-224",
"instructblip-vicuna-7b",
"instructblip-vicuna-13b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "5",
"Rationale": "Text 1 provides a comprehensive list of potential hazards that the skateboarder might encounter, directly addressing the user's prompt. It goes into detail about each hazard, explaining how they could occur and suggesting mea... | The skateboarder might encounter potential hazards such as:
1. **Collision with pedestrians or nearby objects**: The proximity to windows and the trash can could pose a risk for collision.
2. **Unstable or uneven surfaces**: If the ramp or surrounding pavement is uneven or poorly constructed, it could cause loss of balance.
3. **Traffic**: If this area is near or connected to a street, moving vehicles could be a danger.
4. **Loss of momentum**: The skateboarder could fall or fail the stunt if they're unbalanced while coming down the ramp.
5. **Environmental hazards**: Debris on the ramp or sidewalk could act as obstacles, reducing traction and control.
It’s essential that skateboarders wear safety gear, such as helmets and pads, and remain cautious under such conditions. | What potential hazard might the skateboarder encounter on this street? | [
0.8542909026145935,
0.891913890838623,
0.7931522130966187,
0.8491158485412598
] | 0.001244 | 0.980853 | 0.847118 |
svit-complex_reasoning-995 | [
"VisualGLM-6b",
"MMICL-Vicuna-13B",
"llava1.5-7b",
"llava1.5-7b"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response is not helpful as it contains multiple inaccuracies and seems to misunderstand the purpose of the holes in the carton. It discusses UV light exposure and suggests that the eggs' color comes from processing, whic... | The holes in the egg carton underneath each egg serve several purposes:
1. **Stability:** The indents cradle the eggs, preventing them from rolling around and reducing the risk of cracking during transport or storage.
2. **Protection:** The structure keeps each egg separated, absorbing impacts and preventing collisions between eggs that could cause breakage.
3. **Airflow:** The holes allow air to circulate around the eggs, keeping them cool and fresh by minimizing moisture build-up.
4. **Material Efficiency:** The design reduces material usage by not creating a solid base throughout, making it lighter and more cost-effective to produce.
This thoughtful design helps ensure the eggs remain intact and fresh from the farm to your kitchen. | Why are there holes in the carton underneath the eggs? | [
0.6952560544013977,
0.7341292500495911,
0.9166688919067383,
0.9170209169387817
] | 0.010405 | 0.986182 | 0.815769 |
llava-conversation-914 | [
"fuyu-8b",
"instructblip-vicuna-13b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This response is not helpful because it provides an incorrect count of airplanes visible in the image, directly contradicting the visual evidence."
},
"ethical_considerations": {
"Rating": "5",
"Ra... | In the image, there are three airplanes visible. | How many airplanes are visible in the image at the airport? | [
0.11065448820590973,
0.7965751886367798,
0.7965751886367798,
0.874394416809082
] | 0.096024 | 0.957653 | 0.64455 |
LLaVAMed-000000000729 | [
"llava1.5-7b",
"llava1.5-7b",
"instructblip-vicuna-7b",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "The response inaccurately assumes information about the patient's age and gender that cannot be determined from the image. It also inappropriately speculates on the presence of health issues without any evidence from the imag... | This is a screen capture of a 3D imaging software, likely used for medical or dental analysis, and features cross-sectional imagery of a patient's skull. However, based on just the provided image, it is not possible to determine the patient's age or gender definitively. Additional context, clinical data, or more specific and labeled anatomical features would be required for such an assessment. | What can you tell about the patient's age and gender? | [
0.6579716205596924,
0.5833242535591125,
0.7397713661193848,
0.780624508857727
] | 0.005774 | 0.887397 | 0.690423 |
LLaVAR-000000003862 | [
"fuyu-8b",
"instructblip-vicuna-7b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "This text inaccurately attributes the book \"Sweet Home\" to Rob Sherman, which is incorrect and misleading. Therefore, it does not help the user."
},
"ethical_considerations": {
"Rating": "4",
"Ra... | Certainly! The book **"Sweet Home" by Carys Bray** is a collection of short stories that explore the complexities and contradictions of family life, love, and suburban existence. Carys Bray blends her lyrical storytelling with emotional depth, offering tales that are poignant, darkly humorous, and at times heartbreaking.
The stories in the collection delve into themes such as loss, longing, and the pressures of societal expectations, often with a touch of magical realism. The characters in the book grapple with relatable yet extraordinary situations, making the reader question the definition of a "sweet home."
This collection is praised for its ability to capture both the beauty and challenges of human relationships, often delivering a mix of humor and melancholy. It's an introspective and rewarding read for those who enjoy literary fiction and emotional storytelling. Carys Bray is also known for her debut novel **"A Song for Issy Bradley"**, which gained significant acclaim. | Can you tell me more about the book mentioned in the image? | [
0.5565488934516907,
0.5506325960159302,
0.4574734568595886,
0.8099764585494995
] | 0.017142 | 0.942596 | 0.593658 |
LLaVAMed-000000002951 | [
"llava1.5-7b",
"instructblip-vicuna-13b",
"Qwen-VL-Chat",
"gpt-4v"
] | [
{
"annotations": {
"helpfulness": {
"Rating": "1",
"Rationale": "Text 4 is not helpful as it provides an incorrect and vague description of the imaging techniques, suggesting they are used to showcase the brain and spinal cord, which is not clearly supported by the images."
},
... | This figure shows images from **magnetic resonance imaging (MRI)**.
MRI is a medical imaging technique used to visualize internal structures of the body in detail. The images are sagittal views, likely of the lumbar spine. The two images may represent different MRI sequences (e.g., T1-weighted on the left and T2-weighted on the right), which highlight tissues differently (fat vs. fluid content) to aid in diagnosis. | What type of imaging technique is used in this figure? | [
0.661681056022644,
0.58925461769104,
0.6675637364387512,
0.6506971120834351
] | 0.000975 | 0.924487 | 0.642299 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.