id large_stringlengths 10 10 | date date32 | title large_stringlengths 6 202 | author_names large listlengths 1 3.3k | github large_stringlengths 26 132 ⌀ | project_page large_stringlengths 15 247 ⌀ | upvotes int32 0 815 ⌀ | comments int32 0 143 ⌀ | peak int32 0 815 ⌀ | first_day date32 | up_d0 int32 0 524 ⌀ | up_d1 int32 0 695 ⌀ | up_d3 int32 0 683 ⌀ | up_d7 int32 0 695 ⌀ | purged int32 0 558 | up_close int32 0 695 | closed bool 2
classes | week large_stringclasses 177
values | month large_stringdate 2023-05-01 00:00:00 2026-10-01 00:00:00 | early bool 2
classes | rank_day int32 1 109 | n_day int32 2 109 | rank_week int32 1 403 | n_week int32 9 404 | rank_month int32 1 921 | n_month int32 114 946 | summary large_stringlengths 104 1.92k ⌀ | authors large listlengths 1 3.3k ⌀ | org large_stringlengths 2 42 ⌀ | org_fullname large_stringlengths 2 79 ⌀ | org_avatar large_stringlengths 81 108 ⌀ | submitter large_stringlengths 2 31 ⌀ | submitter_fullname large_stringlengths 1 50 ⌀ | submitter_avatar large_stringlengths 45 108 ⌀ | github_stars int64 0 91.9k ⌀ | published_at large_stringlengths 24 24 ⌀ | thumbnail large_stringlengths 77 77 ⌀ | author_participating bool 2
classes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2509.25729 | 2025-10-03 | Controlled Generation for Private Synthetic Text | [
"Zihao Zhao",
"Anjalie Field"
] | https://github.com/zzhao71/Controlled-Generation-for-Private-Synthetic-Text | null | 11 | 2 | 11 | 2025-10-03 | 2 | 2 | 2 | 2 | 0 | 2 | true | 2025-W40 | 2025-10 | false | 45 | 52 | 219 | 275 | 756 | 945 | Text anonymization is essential for responsibly developing and deploying AI
in high-stakes domains such as healthcare, social services, and law. In this
work, we propose a novel methodology for privacy-preserving synthetic text
generation that leverages the principles of de-identification and the Hiding In
Plain Sight ... | [
{
"name": "Zihao Zhao",
"user": "zzhao0104",
"fullname": "Zihao Zhao",
"avatar": "/avatars/b6095891d4260c625fa727927dac9583.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Anjalie Field",
"user": null,
"fullname": null,
"avatar": null,
"status": null... | jhu-clsp | Center for Language and Speech Processing @ JHU | zzhao0104 | Zihao Zhao | /avatars/b6095891d4260c625fa727927dac9583.svg | 5 | 2025-09-30T03:38:36.000Z | true | ||
2411.02657 | 2024-11-06 | Zebra-Llama: A Context-Aware Large Language Model for Democratizing Rare Disease Knowledge | [
"Karthik Soman",
"Andrew Langdon",
"Catalina Villouta",
"Chinmay Agrawal",
"Lashaw Salta",
"Braian Peetoom",
"Gianmarco Bellucci",
"Orion J Buske"
] | https://github.com/karthiksoman/zebra-Llama | null | 6 | 1 | 6 | 2024-11-06 | 3 | 5 | 5 | 5 | 0 | 5 | true | 2024-W45 | 2024-11 | false | 10 | 11 | 70 | 76 | 240 | 285 | Rare diseases present unique challenges in healthcare, often suffering from
delayed diagnosis and fragmented information landscapes. The scarcity of
reliable knowledge in these conditions poses a distinct challenge for Large
Language Models (LLMs) in supporting clinical management and delivering precise
patient informa... | [
{
"name": "Karthik Soman",
"user": "ksoman",
"fullname": "karthik soman",
"avatar": "/avatars/f9eb37ade965a8d99bb50e357498e545.svg",
"status": "extracted_confirmed",
"hidden": false
},
{
"name": "Andrew Langdon",
"user": null,
"fullname": null,
"avatar": null,
"status... | null | null | null | ksoman | karthik soman | /avatars/f9eb37ade965a8d99bb50e357498e545.svg | 17 | 2024-11-04T22:45:52.000Z | true | |
2609.20800 | 2026-09-18 | JEPA-Anything: Learning Predictive Models across Different Worlds | [
"Taoyong Cui",
"Zhongyao Wang",
"Xinyue Xu",
"Weiyang Liu",
"Zhaochen Yu",
"Yuying Zhang",
"Qiang Gao",
"Mengyue Yang",
"Wanli Ouyang",
"Pheng Ann Heng",
"Yingcheng Wu",
"Zhenfei Yin",
"Ling Yang"
] | https://github.com/Gen-Verse/JEPA-Anything | null | 77 | 4 | 77 | 2026-09-18 | 23 | 40 | 58 | 70 | 0 | 70 | true | 2026-W38 | 2026-09 | false | 6 | 24 | 28 | 132 | 109 | 786 | World modeling enables intelligence to anticipate consequences, guide interventions, and learn from interaction. Yet predictive models remain domain-specific: can a common learning principle support world modeling across radically different systems? We introduce JEPA-Anything, a domain-agnostic framework based on ortho... | [
{
"name": "Taoyong Cui",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Zhongyao Wang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Xinyue Xu",
"user": nul... | null | null | null | taesiri | taesiri | 123 | 2026-09-17T00:00:00.000Z | false | ||
2608.13558 | 2026-08-14 | OmniScientist: An Omni-Modal Omni-Discipline AI Scientist | [
"Bobo Li",
"Hao Fei",
"Tianjie Ju",
"Mong-Li Lee",
"Wynne Hsu"
] | https://github.com/Omni-Scientist/OmniScientist | https://omni-scientist.github.io/ | 41 | 2 | 94 | 2026-08-14 | 4 | 6 | 14 | 87 | 53 | 87 | true | 2026-W33 | 2026-08 | false | 5 | 31 | 17 | 155 | 57 | 647 | Recent advances in foundation models have enabled AI scientists to automate increasingly complete research workflows, from hypothesis generation and code execution to manuscript preparation. Yet workflow coverage alone does not provide access to the full evidence on which scientific discovery depends. Existing systems ... | [
{
"name": "Bobo Li",
"user": "BradNLP",
"fullname": "Li Bobo",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/bpuWQVf5a-BFc2dsWnDPS.png",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Hao Fei",
"user": "scofield7419",
"fullname": "Ha... | NationalUniversityofSingapore | National University of Singapore | BradNLP | Li Bobo | 163 | 2026-08-13T00:00:00.000Z | true | |||
2601.04175 | 2026-01-12 | Legal Alignment for Safe and Ethical AI | [
"Noam Kolt",
"Nicholas Caputo",
"Jack Boeglin",
"Cullen O'Keefe",
"Rishi Bommasani",
"Stephen Casper",
"Mariano-Florentino Cuéllar",
"Noah Feldman",
"Iason Gabriel",
"Gillian K. Hadfield",
"Lewis Hammond",
"Peter Henderson",
"Atoosa Kasirzadeh",
"Seth Lazar",
"Anka Reuel",
"Kevin L. We... | null | https://www.legal-alignment.ai/ | 5 | 3 | 5 | 2026-01-12 | 0 | 3 | 3 | 3 | 0 | 3 | true | 2026-W03 | 2026-01 | false | 23 | 31 | 121 | 162 | 446 | 573 | Alignment of artificial intelligence (AI) encompasses the normative problem of specifying how AI systems should act and the technical problem of ensuring AI systems comply with those specifications. To date, AI alignment has generally overlooked an important source of knowledge and practice for grappling with these pro... | [
{
"name": "Noam Kolt",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Nicholas Caputo",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Jack Boeglin",
"user": ... | null | null | null | noamkolt | Noam Kolt | /avatars/63d59cbe46a7c8ce02d93f1370a0af2b.svg | null | 2026-01-07T18:42:04.000Z | false | |
2305.17098 | 2023-05-29 | ControlVideo: Adding Conditional Control for One Shot Text-to-Video Editing | [
"Min Zhao",
"Rongzhen Wang",
"Fan Bao",
"Chongxuan Li",
"Jun Zhu"
] | https://github.com/thu-ml/controlvideo | null | 5 | 3 | 5 | 2024-03-12 | null | null | null | null | 0 | 5 | true | 2023-W22 | 2023-05 | true | 2 | 21 | 20 | 86 | 51 | 258 | In this paper, we present ControlVideo, a novel method for text-driven video
editing. Leveraging the capabilities of text-to-image diffusion models and
ControlNet, ControlVideo aims to enhance the fidelity and temporal consistency
of videos that align with a given text while preserving the structure of the
source video... | [
{
"name": "Min Zhao",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Rongzhen Wang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Fan Bao",
"user": "baofff"... | null | null | null | akhaliq | AK | 230 | 2023-05-26T17:13:55.000Z | false | ||
2605.08354 | 2026-05-12 | Auto-Rubric as Reward: From Implicit Preferences to Explicit Multimodal Generative Criteria | [
"Juanxi Tian",
"Fengyuan Liu",
"Jiaming Han",
"Yilei Jiang",
"Yongliang Wu",
"Yesheng Liu",
"Haodong Li",
"Furong Xu",
"Wanhua Li"
] | https://github.com/OpenEnvision/AutoRubric-as-Reward | https://openenvision.github.io/AutoRubric-as-Reward/ | 21 | 2 | 23 | 2026-05-12 | 20 | 21 | 22 | 23 | 2 | 23 | true | 2026-W20 | 2026-05 | false | 13 | 68 | 64 | 305 | 242 | 939 | Aligning multimodal generative models with human preferences demands reward signals that respect the compositional, multi-dimensional structure of human judgment. Prevailing RLHF approaches reduce this structure to scalar or pairwise labels, collapsing nuanced preferences into opaque parametric proxies and exposing vul... | [
{
"name": "Juanxi Tian",
"user": "Juanxi",
"fullname": "Juanxi Tian",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/670880950e79a8b46f7ff9dd/hA1TLhwlQblkFsq8wLrkB.jpeg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Fengyuan Liu",
"user": "Ferr... | OpenEnvisionLab | OpenEnvision | Juanxi | Juanxi Tian | 57 | 2026-05-08T00:00:00.000Z | true | |||
2512.14699 | 2025-12-17 | MemFlow: Flowing Adaptive Memory for Consistent and Efficient Long Video Narratives | [
"Sihui Ji",
"Xi Chen",
"Shuai Yang",
"Xin Tao",
"Pengfei Wan",
"Hengshuang Zhao"
] | https://github.com/KlingTeam/MemFlow | null | 29 | 1 | 29 | 2025-12-17 | 14 | 20 | 25 | 26 | 0 | 26 | true | 2025-W51 | 2025-12 | false | 9 | 39 | 39 | 170 | 148 | 633 | The core challenge for streaming video generation is maintaining the content consistency in long context, which poses high requirement for the memory design. Most existing solutions maintain the memory by compressing historical frames with predefined strategies. However, different to-generate video chunks should refer ... | [
{
"name": "Sihui Ji",
"user": "zjuJish",
"fullname": "Sihui Ji",
"avatar": "/avatars/19af37ffd7626ebc58b387a34a8f98d7.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Xi Chen",
"user": "xichenhku",
"fullname": "xichen",
"avatar": "https://cdn-avatars.hugg... | null | null | null | taesiri | taesiri | 218 | 2025-12-16T18:59:59.000Z | false | ||
2508.11548 | 2025-08-20 | Copyright Protection for Large Language Models: A Survey of Methods, Challenges, and Trends | [
"Zhenhua Xu",
"Xubin Yue",
"Zhebo Wang",
"Qichen Liu",
"Xixiang Zhao",
"Jingxuan Zhang",
"Wenjun Zeng",
"Wengpeng Xing",
"Dezhang Kong",
"Changting Lin",
"Meng Han"
] | https://github.com/Xuzhenhua55/awesome-llm-copyright-protection | https://xuzhenhua55.github.io/awesome-llm-copyright-protection/index.html | 4 | 2 | 5 | 2025-08-20 | 5 | 5 | 5 | 5 | 1 | 5 | true | 2025-W34 | 2025-08 | false | 16 | 28 | 68 | 95 | 320 | 452 | Copyright protection for large language models is of critical importance,
given their substantial development costs, proprietary value, and potential for
misuse. Existing surveys have predominantly focused on techniques for tracing
LLM-generated content-namely, text watermarking-while a systematic exploration
of method... | [
{
"name": "Zhenhua Xu",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Xubin Yue",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Zhebo Wang",
"user": "Breyna... | null | null | null | BreynaldDva | Zhebo Wang | 52 | 2025-08-15T15:50:20.000Z | true | ||
2507.05791 | 2025-07-09 | GTA1: GUI Test-time Scaling Agent | [
"Yan Yang",
"Dongxu Li",
"Yutong Dai",
"Yuhao Yang",
"Ziyang Luo",
"Zirui Zhao",
"Zhiyuan Hu",
"Junzhe Huang",
"Amrita Saha",
"Zeyuan Chen",
"Ran Xu",
"Liyuan Pan",
"Caiming Xiong",
"Junnan Li"
] | https://github.com/Yan98/GTA1 | null | 27 | 1 | 27 | 2025-07-09 | 15 | 21 | 24 | 24 | 0 | 24 | true | 2025-W28 | 2025-07 | false | 9 | 28 | 35 | 99 | 133 | 383 | Graphical user interface (GUI) agents autonomously operate across platforms
(e.g., Linux) to complete tasks by interacting with visual elements.
Specifically, a user instruction is decomposed into a sequence of action
proposals, each corresponding to an interaction with the GUI. After each
action, the agent observes th... | [
{
"name": "Yan Yang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Dongxu Li",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Yutong Dai",
"user": null,
... | Salesforce | Salesforce AI Research | HelloKKMe | Yan Yang | /avatars/8633309cae0bbce566a2540e639e5688.svg | 130 | 2025-07-08T08:52:18.000Z | false | ||
2507.16880 | 2025-07-24 | Finding Dori: Memorization in Text-to-Image Diffusion Models Is Less Local Than Assumed | [
"Antoni Kowalczuk",
"Dominik Hintersdorf",
"Lukas Struppek",
"Kristian Kersting",
"Adam Dziedzic",
"Franziska Boenisch"
] | null | null | 7 | 1 | 7 | 2025-07-24 | 3 | 6 | 6 | 6 | 0 | 6 | true | 2025-W30 | 2025-07 | false | 11 | 11 | 68 | 86 | 285 | 383 | Text-to-image diffusion models (DMs) have achieved remarkable success in
image generation. However, concerns about data privacy and intellectual
property remain due to their potential to inadvertently memorize and replicate
training data. Recent mitigation efforts have focused on identifying and
pruning weights respons... | [
{
"name": "Antoni Kowalczuk",
"user": "antoniaaa",
"fullname": "Antoni Kowalczuk",
"avatar": "/avatars/5e0dbe8e122b3106b7799e226cb9a9c7.svg",
"status": "admin_assigned",
"hidden": false
},
{
"name": "Dominik Hintersdorf",
"user": "D0miH",
"fullname": "Dominik Hintersdorf",
... | null | null | null | lukas-struppek | Lukas Struppek | null | 2025-07-22T15:02:38.000Z | true | ||
2509.06733 | 2025-09-09 | Reinforcement Learning Foundations for Deep Research Systems: A Survey | [
"Wenjun Li",
"Zhi Chen",
"Jingru Lin",
"Hannan Cao",
"Wei Han",
"Sheng Liang",
"Zhi Zhang",
"Kuicai Dong",
"Dexun Li",
"Chen Zhang",
"Yong Liu"
] | https://github.com/wenjunli-0/deepresearch-survey | null | 32 | 2 | 32 | 2025-09-09 | 21 | 26 | 28 | 31 | 0 | 31 | true | 2025-W37 | 2025-09 | false | 5 | 25 | 26 | 92 | 108 | 536 | Deep research systems, agentic AI that solve complex, multi-step tasks by
coordinating reasoning, search across the open web and user files, and tool
use, are moving toward hierarchical deployments with a Planner, Coordinator,
and Executors. In practice, training entire stacks end-to-end remains
impractical, so most wo... | [
{
"name": "Wenjun Li",
"user": "wenjun-li",
"fullname": "wenjun",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/6622245224f3842a31f7c58a/EcvwtHamYwMdzRT9pevdE.jpeg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Zhi Chen",
"user": "hellochenzhi... | null | null | null | wenjun-li | wenjun | 48 | 2025-09-08T14:27:23.000Z | true | ||
2503.15855 | 2025-03-21 | VideoRFSplat: Direct Scene-Level Text-to-3D Gaussian Splatting Generation with Flexible Pose and Multi-View Joint Modeling | [
"Hyojun Go",
"Byeongjun Park",
"Hyelin Nam",
"Byung-Hoon Kim",
"Hyungjin Chung",
"Changick Kim"
] | https://github.com/gohyojun15/VideoRFSplat | https://gohyojun15.github.io/VideoRFSplat/ | 5 | 2 | 5 | 2025-03-21 | 2 | 3 | 3 | 3 | 0 | 3 | true | 2025-W12 | 2025-03 | false | 43 | 48 | 139 | 155 | 513 | 611 | We propose VideoRFSplat, a direct text-to-3D model leveraging a video
generation model to generate realistic 3D Gaussian Splatting (3DGS) for
unbounded real-world scenes. To generate diverse camera poses and unbounded
spatial extent of real-world scenes, while ensuring generalization to arbitrary
text prompts, previous... | [
{
"name": "Hyojun Go",
"user": "HJGO",
"fullname": "Hyojun GO",
"avatar": "/avatars/b0dcd8ad795b1e666ee247b2ac024d53.svg",
"status": "admin_assigned",
"hidden": false
},
{
"name": "Byeongjun Park",
"user": "byeongjun-park",
"fullname": "Byeongjun Park",
"avatar": "/avatar... | everex | EverEx | HJGO | Hyojun GO | /avatars/b0dcd8ad795b1e666ee247b2ac024d53.svg | 23 | 2025-03-20T05:26:09.000Z | true | ||
2506.00469 | 2025-06-03 | Massively Multilingual Adaptation of Large Language Models Using Bilingual Translation Data | [
"Shaoxiong Ji",
"Zihao Li",
"Jaakko Paavola",
"Indraneil Paul",
"Hengyu Luo",
"Jörg Tiedemann"
] | https://github.com/MaLA-LM/emma-500 | https://mala-lm.github.io/emma-500-gen2.html | 4 | 2 | 4 | 2025-06-03 | 1 | 2 | 2 | 2 | 0 | 2 | true | 2025-W23 | 2025-06 | false | 51 | 62 | 206 | 250 | 591 | 679 | This paper investigates a critical design decision in the practice of
massively multilingual continual pre-training -- the inclusion of parallel
data. Specifically, we study the impact of bilingual translation data for
massively multilingual language adaptation of the Llama3 family of models to
500 languages. To this e... | [
{
"name": "Shaoxiong Ji",
"user": "jisx",
"fullname": "Shaoxiong",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/617a92e16f37340367d5d791/omgyzmaF90KBLa3YgFxhS.png",
"status": "claimed_verified",
"hidden": true
},
{
"name": "Zihao Li",
"user": "Zihao-Li",
... | null | null | null | jisx | Shaoxiong | 3 | 2025-05-31T08:37:17.000Z | true | ||
2409.12191 | 2024-09-19 | Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution | [
"Peng Wang",
"Shuai Bai",
"Sinan Tan",
"Shijie Wang",
"Zhihao Fan",
"Jinze Bai",
"Keqin Chen",
"Xuejing Liu",
"Jialin Wang",
"Wenbin Ge",
"Yang Fan",
"Kai Dang",
"Mengfei Du",
"Xuancheng Ren",
"Rui Men",
"Dayiheng Liu",
"Chang Zhou",
"Jingren Zhou",
"Junyang Lin"
] | https://github.com/qwenlm/qwen2-vl | null | 80 | 6 | 80 | 2024-09-19 | 39 | 54 | 55 | 63 | 1 | 63 | true | 2024-W38 | 2024-09 | false | 2 | 17 | 4 | 70 | 13 | 254 | We present the Qwen2-VL Series, an advanced upgrade of the previous Qwen-VL
models that redefines the conventional predetermined-resolution approach in
visual processing. Qwen2-VL introduces the Naive Dynamic Resolution mechanism,
which enables the model to dynamically process images of varying resolutions
into differe... | [
{
"name": "Peng Wang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Shuai Bai",
"user": "ShuaiBai623",
"fullname": "shuai bai",
"avatar": "/avatars/dec4bbee4a82b773fc58dfc2dce9dbeb.svg",
"status": "admin_assigned",
... | null | null | null | akhaliq | AK | 20,034 | 2024-09-18T17:59:32.000Z | true | ||
2511.11113 | 2025-11-19 | VIDEOP2R: Video Understanding from Perception to Reasoning | [
"Yifan Jiang",
"Yueying Wang",
"Rui Zhao",
"Toufiq Parag",
"Zhimin Chen",
"Zhenyu Liao",
"Jayakrishnan Unnikrishnan"
] | null | ERROR: type should be large_string, got "https://videop2r.\ngithub.io/videop2r/" | 113 | 5 | 113 | 2025-11-20 | null | 98 | 104 | 107 | 0 | 107 | true | 2025-W47 | 2025-11 | false | 1 | 20 | 5 | 105 | 14 | 412 | Reinforcement fine-tuning (RFT), a two-stage framework consisting of supervised fine-tuning (SFT) and reinforcement learning (RL) has shown promising results on improving reasoning ability of large language models (LLMs). Yet extending RFT to large video language models (LVLMs) remains challenging. We propose VideoP2R,... | [
{
"name": "Yifan Jiang",
"user": "YifanJ",
"fullname": "Yifan Jiang",
"avatar": "/avatars/61014a3fd45f74ce541bdbe53929e233.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Yueying Wang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
... | amazon | Amazon | YifanJ | Yifan Jiang | /avatars/61014a3fd45f74ce541bdbe53929e233.svg | null | 2025-11-14T09:42:42.000Z | true | ||
2511.19418 | 2025-11-25 | Chain-of-Visual-Thought: Teaching VLMs to See and Think Better with Continuous Visual Tokens | [
"Yiming Qin",
"Bomin Wei",
"Jiaxin Ge",
"Konstantinos Kallidromitis",
"Stephanie Fu",
"Trevor Darrell",
"Xudong Wang"
] | https://github.com/Wakals/CoVT | https://wakalsprojectpage.github.io/covt-website/ | 29 | 4 | 29 | 2025-11-25 | 17 | 20 | 23 | 25 | 0 | 25 | true | 2025-W48 | 2025-11 | false | 9 | 31 | 29 | 115 | 119 | 412 | Vision-Language Models (VLMs) excel at reasoning in linguistic space but struggle with perceptual understanding that requires dense visual perception, e.g., spatial reasoning and geometric awareness. This limitation stems from the fact that current VLMs have limited mechanisms to capture dense visual information across... | [
{
"name": "Yiming Qin",
"user": "Wakals",
"fullname": "YM Qin",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/6527fe894126af677af1541b/uTjqrkJ92NcstRskuPvhj.png",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Bomin Wei",
"user": null,
"full... | UCBerkeley | University of California, Berkeley | taesiri | taesiri | 412 | 2025-11-24T18:55:19.000Z | false | |||
2503.01370 | 2025-03-04 | Kiss3DGen: Repurposing Image Diffusion Models for 3D Asset Generation | [
"Jiantao Lin",
"Xin Yang",
"Meixi Chen",
"Yingjie Xu",
"Dongyu Yan",
"Leyi Wu",
"Xinli Xu",
"Lie XU",
"Shunsi Zhang",
"Ying-Cong Chen"
] | https://github.com/EnVision-Research/Kiss3DGen | https://ltt-o.github.io/Kiss3dgen.github.io/ | 15 | 2 | 15 | 2025-03-04 | 7 | 7 | 7 | 9 | 0 | 9 | true | 2025-W10 | 2025-03 | false | 15 | 28 | 58 | 114 | 313 | 611 | Diffusion models have achieved great success in generating 2D images.
However, the quality and generalizability of 3D content generation remain
limited. State-of-the-art methods often require large-scale 3D assets for
training, which are challenging to collect. In this work, we introduce
Kiss3DGen (Keep It Simple and S... | [
{
"name": "Jiantao Lin",
"user": "LTT",
"fullname": "JIANTAO LIN",
"avatar": "/avatars/c1922acfda2e6d2fe7b03194a404eb10.svg",
"status": "admin_assigned",
"hidden": false
},
{
"name": "Xin Yang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidd... | null | null | null | LTT | JIANTAO LIN | /avatars/c1922acfda2e6d2fe7b03194a404eb10.svg | 298 | 2025-03-03T10:07:19.000Z | true | |
2511.07384 | 2025-11-11 | Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrence | [
"Sean McLeish",
"Ang Li",
"John Kirchenbauer",
"Dayal Singh Kalra",
"Brian R. Bartoldson",
"Bhavya Kailkhura",
"Avi Schwarzschild",
"Jonas Geiping",
"Tom Goldstein",
"Micah Goldblum"
] | https://github.com/mcleish7/retrofitting-recurrence | null | 21 | 2 | 21 | 2025-11-11 | 8 | 10 | 13 | 15 | 0 | 15 | true | 2025-W46 | 2025-11 | false | 12 | 31 | 34 | 86 | 166 | 412 | Recent advances in depth-recurrent language models show that recurrence can
decouple train-time compute and parameter count from test-time compute. In this
work, we study how to convert existing pretrained non-recurrent language models
into depth-recurrent models. We find that using a curriculum of recurrences to
incre... | [
{
"name": "Sean McLeish",
"user": "smcleish",
"fullname": "Sean McLeish",
"avatar": "/avatars/257085f01c439d7c84787a4e6d085b3d.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Ang Li",
"user": "leonli66",
"fullname": "Leon Li",
"avatar": "https://cdn-avat... | tomg-group-umd | Tom Goldstein's Lab at University of Maryland, College Park | smcleish | Sean McLeish | /avatars/257085f01c439d7c84787a4e6d085b3d.svg | 70 | 2025-11-10T18:43:07.000Z | true | ||
2605.04615 | 2026-05-11 | Beyond Retrieval: A Multitask Benchmark and Model for Code Search | [
"Siqiao Xue",
"Zihan Liao",
"Jin Qin",
"Ziyin Zhang",
"Yixiang Mu",
"Fan Zhou",
"Hang Yu"
] | https://github.com/hq-bench/coreb | https://hq-bench.github.io/coreb-page/ | 19 | 2 | 24 | 2026-05-11 | 22 | 22 | 23 | 23 | 5 | 23 | true | 2026-W20 | 2026-05 | false | 11 | 57 | 64 | 305 | 242 | 939 | Code search has usually been evaluated as first-stage retrieval, even though production systems rely on broader pipelines with reranking and developer-style queries. Existing benchmarks also suffer from data contamination, label noise, and degenerate binary relevance. In this paper, we introduce CoREB, a contamination-... | [
{
"name": "Siqiao Xue",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Zihan Liao",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Jin Qin",
"user": null,
... | hq-bench | high-quality llm benchmarks | https://www.gravatar.com/avatar/3a6308fe52f03e3083fbd5b657d9268b?d=retro&size=100 | Geralt-Targaryen | Ziyin Zhang | 2 | 2026-05-06T00:00:00.000Z | true | ||
2406.15339 | 2024-06-26 | Image Conductor: Precision Control for Interactive Video Synthesis | [
"Yaowei Li",
"Xintao Wang",
"Zhaoyang Zhang",
"Zhouxia Wang",
"Ziyang Yuan",
"Liangbin Xie",
"Yuexian Zou",
"Ying Shan"
] | null | null | 9 | 3 | 9 | 2024-06-26 | 6 | 8 | 8 | 8 | 0 | 8 | true | 2024-W26 | 2024-06 | false | 15 | 20 | 67 | 103 | 228 | 346 | Filmmaking and animation production often require sophisticated techniques
for coordinating camera transitions and object movements, typically involving
labor-intensive real-world capturing. Despite advancements in generative AI for
video creation, achieving precise control over motion for interactive video
asset gener... | [
{
"name": "Yaowei Li",
"user": "Yw22",
"fullname": "Yaowei Li",
"avatar": "/avatars/041ad5abf9be42e336938f51ebb8746c.svg",
"status": "admin_assigned",
"hidden": false
},
{
"name": "Xintao Wang",
"user": "Xintao",
"fullname": "Xintao Wang",
"avatar": "https://cdn-avatars.h... | null | null | null | Liangbin | Liangbin Xie | /avatars/c58d228dbdf4f1edb5aa4e635dacb51d.svg | null | 2024-06-21T17:55:05.000Z | true | |
2502.16069 | 2025-02-26 | Curie: Toward Rigorous and Automated Scientific Experimentation with AI Agents | [
"Patrick Tser Jern Kon",
"Jiachen Liu",
"Qiuyi Ding",
"Yiming Qiu",
"Zhenning Yang",
"Yibo Huang",
"Jayanth Srinivasa",
"Myungjin Lee",
"Mosharaf Chowdhury",
"Ang Chen"
] | https://github.com/Just-Curieous/Curie | null | 20 | 5 | 20 | 2025-02-26 | 9 | 16 | 16 | 17 | 0 | 17 | true | 2025-W09 | 2025-02 | false | 7 | 22 | 46 | 134 | 179 | 502 | Scientific experimentation, a cornerstone of human progress, demands rigor in
reliability, methodical control, and interpretability to yield meaningful
results. Despite the growing capabilities of large language models (LLMs) in
automating different aspects of the scientific process, automating rigorous
experimentation... | [
{
"name": "Patrick Tser Jern Kon",
"user": "patkon",
"fullname": "Patrick Kon",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/64b7111e17681d64b19cf95e/VHPfCUl1nBS3OMMVi96CR.jpeg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Jiachen Liu",
"use... | null | null | null | AmberLJC | Amber | null | 2025-02-22T03:58:19.000Z | false | ||
2606.16519 | 2026-06-16 | BadWorld: Adversarial Attacks on World Models | [
"Linghui Shen",
"Mingyue Cui",
"Xingyi Yang"
] | https://github.com/LinghuiiShen/BadWorld | https://linghuiishen.github.io/BadWorld/ | 16 | 1 | 18 | 2026-06-16 | 14 | 14 | 15 | 18 | 2 | 18 | true | 2026-W25 | 2026-06 | false | 11 | 42 | 54 | 192 | 280 | 946 | Visual world models (VWMs) synthesize interactive, action-conditioned rollouts from a single context image. However, it remains an open question how robust these models are to adversarial perturbations. Standard adversarial attacks fail to assess this vulnerability because attackers lack ground-truth future videos and ... | [
{
"name": "Linghui Shen",
"user": "LinghuiShen",
"fullname": "Shen Linghui",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/2liMvKYwnE2xNGmpE_GwW.png",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Mingyue Cui",
"user": "Mingyueee",
... | PolyUHK | The Hong Kong Polytechnic University | adamdad | Xingyi Yang | 19 | 2026-06-15T00:00:00.000Z | true | |||
2405.13817 | 2024-05-24 | Thermodynamic Natural Gradient Descent | [
"Kaelan Donatella",
"Samuel Duffield",
"Maxwell Aifer",
"Denis Melanson",
"Gavin Crooks",
"Patrick J. Coles"
] | null | null | 15 | 1 | 17 | 2024-05-24 | 8 | 9 | 10 | 12 | 3 | 12 | true | 2024-W21 | 2024-05 | false | 7 | 17 | 19 | 36 | 87 | 155 | Second-order training methods have better convergence properties than
gradient descent but are rarely used in practice for large-scale training due
to their computational overhead. This can be viewed as a hardware limitation
(imposed by digital computers). Here we show that natural gradient descent
(NGD), a second-orde... | [
{
"name": "Kaelan Donatella",
"user": "KaelanDt",
"fullname": "Kaelan Donatella",
"avatar": "/avatars/fb05a807a1f5f35c143082adc39dae12.svg",
"status": "extracted_confirmed",
"hidden": false
},
{
"name": "Samuel Duffield",
"user": null,
"fullname": null,
"avatar": null,
... | null | null | null | akhaliq | AK | null | 2024-05-22T16:47:03.000Z | false | ||
2510.04533 | 2025-10-13 | TAG:Tangential Amplifying Guidance for Hallucination-Resistant Diffusion Sampling | [
"Hyunmin Cho",
"Donghoon Ahn",
"Susung Hong",
"Jee Eun Kim",
"Seungryong Kim",
"Kyong Hwan Jin"
] | https://github.com/hyeon-cho/Tangential-Amplifying-Guidance | https://hyeon-cho.github.io/TAG/ | 39 | 6 | 48 | 2025-10-13 | 38 | 44 | 45 | 46 | 9 | 46 | true | 2025-W42 | 2025-10 | false | 5 | 46 | 25 | 244 | 102 | 945 | Recent diffusion models achieve the state-of-the-art performance in image
generation, but often suffer from semantic inconsistencies or hallucinations.
While various inference-time guidance methods can enhance generation, they
often operate indirectly by relying on external signals or architectural
modifications, which... | [
{
"name": "Hyunmin Cho",
"user": "hyeoncho01",
"fullname": "Hyunmin Cho",
"avatar": "/avatars/904d243f2fad99341f11795e93788993.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Donghoon Ahn",
"user": null,
"fullname": null,
"avatar": null,
"status": nu... | null | null | null | hyeoncho01 | Hyunmin Cho | /avatars/904d243f2fad99341f11795e93788993.svg | 45 | 2025-10-06T06:53:29.000Z | true | |
2403.13802 | 2024-03-21 | ZigMa: Zigzag Mamba Diffusion Model | [
"Vincent Tao Hu",
"Stefan Andreas Baumann",
"Ming Gui",
"Olga Grebenkova",
"Pingchuan Ma",
"Johannes Fischer",
"Bjorn Ommer"
] | https://github.com/CompVis/zigma | null | 18 | 2 | 18 | 2024-03-21 | 6 | 9 | 12 | 14 | 0 | 14 | true | 2024-W12 | 2024-03 | false | 9 | 18 | 29 | 67 | 104 | 218 | The diffusion model has long been plagued by scalability and quadratic
complexity issues, especially within transformer-based structures. In this
study, we aim to leverage the long sequence modeling capability of a
State-Space Model called Mamba to extend its applicability to visual data
generation. Firstly, we identif... | [
{
"name": "Vincent Tao Hu",
"user": "taohu",
"fullname": "Vincent Tao Hu",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/63dcd7ac22cc06e76a8484ce/4O8NcevQ0ZWhJJ1r4Pbam.jpeg",
"status": "admin_assigned",
"hidden": false
},
{
"name": "Stefan Andreas Baumann",
... | null | null | null | akhaliq | AK | 351 | 2024-03-20T17:59:14.000Z | false | ||
2504.01201 | 2025-04-03 | Medical large language models are easily distracted | [
"Krithik Vishwanath",
"Anton Alyakin",
"Daniel Alexander Alber",
"Jin Vivian Lee",
"Douglas Kondziolka",
"Eric Karl Oermann"
] | https://github.com/nyuolab/MedDistractQA | null | 3 | 2 | 3 | 2025-04-03 | 1 | 2 | 3 | 3 | 1 | 3 | true | 2025-W14 | 2025-04 | false | 20 | 22 | 115 | 124 | 419 | 456 | Large language models (LLMs) have the potential to transform medicine, but
real-world clinical scenarios contain extraneous information that can hinder
performance. The rise of assistive technologies like ambient dictation, which
automatically generates draft notes from live patient encounters, has the
potential to int... | [
{
"name": "Krithik Vishwanath",
"user": "KrithikV",
"fullname": "Krithik Vishwanath",
"avatar": "/avatars/2c4791516aae0b20a9c9ce0542dc966e.svg",
"status": "extracted_pending",
"hidden": false
},
{
"name": "Anton Alyakin",
"user": "alyakin314",
"fullname": "Anton",
"avatar... | NYU-OLAB | NYU Langone Health OLAB | KrithikV | Krithik Vishwanath | /avatars/2c4791516aae0b20a9c9ce0542dc966e.svg | 6 | 2025-04-01T21:34:01.000Z | true | ||
2602.23165 | 2026-02-27 | DyaDiT: A Multi-Modal Diffusion Transformer for Socially Favorable Dyadic Gesture Generation | [
"Yichen Peng",
"Jyun-Ting Song",
"Siyeol Jung",
"Ruofan Liu",
"Haiyang Liu",
"Xuangeng Chu",
"Ruicong Liu",
"Erwin Wu",
"Hideki Koike",
"Kris Kitani"
] | null | null | 3 | 1 | 3 | 2026-02-27 | 1 | 1 | 1 | 1 | 0 | 1 | true | 2026-W09 | 2026-02 | false | 25 | 28 | 110 | 134 | 675 | 785 | Generating realistic conversational gestures are essential for achieving natural, socially engaging interactions with digital humans. However, existing methods typically map a single audio stream to a single speaker's motion, without considering social context or modeling the mutual dynamics between two people engaging... | [
{
"name": "Yichen Peng",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Jyun-Ting Song",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Siyeol Jung",
"user": ... | null | null | null | taesiri | taesiri | null | 2026-02-26T16:30:07.000Z | false | ||
2602.03295 | 2026-02-04 | POP: Prefill-Only Pruning for Efficient Large Model Inference | [
"Junhui He",
"Zhihui Fu",
"Jun Wang",
"Qingan Li"
] | null | null | 4 | 3 | 4 | 2026-02-04 | 3 | 4 | 4 | 4 | 0 | 4 | true | 2026-W06 | 2026-02 | false | 38 | 53 | 186 | 267 | 511 | 785 | Large Language Models (LLMs) and Vision-Language Models (VLMs) have demonstrated remarkable capabilities. However, their deployment is hindered by significant computational costs. Existing structured pruning methods, while hardware-efficient, often suffer from significant accuracy degradation. In this paper, we argue t... | [
{
"name": "Junhui He",
"user": "Junhuihe",
"fullname": "Junhui He",
"avatar": "/avatars/a8ca4d00c4017189f04905da425c8697.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Zhihui Fu",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"... | null | null | null | Junhuihe | Junhui He | /avatars/a8ca4d00c4017189f04905da425c8697.svg | null | 2026-02-03T09:22:26.000Z | true | |
2609.17488 | 2026-09-16 | LimiX-2: A Contextual Mechanism Network Towards General Structured-Data Intelligence | [
"Xingxuan Zhang",
"Gang Ren",
"Hao Yuan",
"Hao Zou",
"Hongze Tan",
"Hui Wang",
"Jianhao Song",
"Jiansheng Li",
"Jiayao Zhang",
"Jinghan Zhang",
"Kaifang Li",
"Lang Mo",
"Li Mao",
"Mingchao Hao",
"Nuo Xu",
"Rui Ding",
"Ruiji Zhang",
"Shuyang Li",
"Siyu Mei",
"Tianyang Zhang",
... | https://github.com/limix-ldm-ai/LimiX | https://www.limix.ai/ | 815 | 4 | 815 | 2026-09-17 | null | 104 | 279 | 660 | 0 | 660 | true | 2026-W38 | 2026-09 | false | 1 | 26 | 2 | 132 | 2 | 786 | We introduce LimiX-2, a new model in the LimiX family, developed through model and data scaling guided by our previously established scaling laws. LimiX-2 adopts the Contextual Mechanism Networks (CMNs) paradigm and is pretrained with Context-Conditional Masked Modeling (CCMM). CMNs shifts the organizing principle of i... | [
{
"name": "Xingxuan Zhang",
"user": "xuange",
"fullname": "Xingxuan Zhang",
"avatar": "/avatars/622e8ca69f1867be7932d03b91eaf6e7.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Gang Ren",
"user": null,
"fullname": null,
"avatar": null,
"status": null... | stable-ai | Stable AI | xuange | Xingxuan Zhang | /avatars/622e8ca69f1867be7932d03b91eaf6e7.svg | 4,367 | 2026-09-15T00:00:00.000Z | true | ||
2506.20670 | 2025-06-25 | MMSearch-R1: Incentivizing LMMs to Search | [
"Jinming Wu",
"Zihao Deng",
"Wei Li",
"Yiding Liu",
"Bo You",
"Bo Li",
"Zejun Ma",
"Ziwei Liu"
] | https://github.com/EvolvingLMMs-Lab/multimodal-search-r1 | null | 46 | 3 | 65 | 2025-06-26 | null | 19 | 40 | 56 | 19 | 56 | true | 2025-W26 | 2025-06 | false | 3 | 23 | 9 | 115 | 36 | 679 | Robust deployment of large multimodal models (LMMs) in real-world scenarios
requires access to external knowledge sources, given the complexity and dynamic
nature of real-world information. Existing approaches such as
retrieval-augmented generation (RAG) and prompt engineered search agents rely
on rigid pipelines, ofte... | [
{
"name": "Jinming Wu",
"user": "kimingng",
"fullname": "Jinming Wu",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/652fbe8cb2acab0b82f855a6/lVpzeEoFRQ6dnGAoNS9b3.jpeg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Zihao Deng",
"user": null,
... | null | null | null | kimingng | Jinming Wu | 488 | 2025-06-25T17:59:42.000Z | true | ||
2509.22630 | 2025-09-29 | StateX: Enhancing RNN Recall via Post-training State Expansion | [
"Xingyu Shen",
"Yingfa Chen",
"Zhen Leng Thai",
"Xu Han",
"Zhiyuan Liu",
"Maosong Sun"
] | https://github.com/thunlp/StateX | null | 4 | 2 | 4 | 2025-09-29 | 1 | 1 | 1 | 2 | 0 | 2 | true | 2025-W40 | 2025-09 | false | 37 | 44 | 219 | 275 | 445 | 536 | While Transformer-based models have demonstrated remarkable language modeling
performance, their high complexities result in high costs when processing long
contexts. In contrast, recurrent neural networks (RNNs) such as linear
attention and state space models have gained popularity due to their constant
per-token comp... | [
{
"name": "Xingyu Shen",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Yingfa Chen",
"user": "chen-yingfa",
"fullname": "Yingfa Chen",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/6144e4667f2544bb45078... | null | null | null | chen-yingfa | Yingfa Chen | 3 | 2025-09-26T17:55:22.000Z | true | ||
2305.16958 | 2023-05-29 | MixCE: Training Autoregressive Language Models by Mixing Forward and Reverse Cross-Entropies | [
"Shiyue Zhang",
"Shijie Wu",
"Ozan Irsoy",
"Steven Lu",
"Mohit Bansal",
"Mark Dredze",
"David Rosenberg"
] | https://github.com/bloomberg/mixce-acl2023 | null | 2 | 0 | 2 | 2024-03-12 | null | null | null | null | 0 | 2 | true | 2023-W22 | 2023-05 | true | 11 | 21 | 52 | 86 | 124 | 258 | Autoregressive language models are trained by minimizing the cross-entropy of
the model distribution Q relative to the data distribution P -- that is,
minimizing the forward cross-entropy, which is equivalent to maximum likelihood
estimation (MLE). We have observed that models trained in this way may
"over-generalize",... | [
{
"name": "Shiyue Zhang",
"user": "shiyue",
"fullname": "Shiyue Zhang",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/1632426687077-613a776517297f3c0bfd7b41.jpeg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Shijie Wu",
"user": "shijie-wu",
... | null | null | null | akhaliq | AK | 20 | 2023-05-26T14:14:51.000Z | false | ||
2608.21486 | 2026-08-25 | EXPL-FR: Explaining Face Recognition Models via Vision-Language Alignment | [
"Guray Ozgur",
"Mustafa Efe Tamyapar",
"Naser Damer",
"Fadi Boutros"
] | https://github.com/gurayozgur/EXPL-FR | https://expl-fr.github.io/ | 0 | 2 | 1 | 2026-08-25 | 1 | 1 | 1 | 1 | 1 | 1 | true | 2026-W35 | 2026-08 | false | 34 | 35 | 135 | 139 | 638 | 647 | Deep face recognition (FR) models reach near-saturated accuracy but remain opaque: a practitioner cannot ask which semantic attributes a similarity score relied upon. EXPL-FR answers this inside the FR model's own embedding space. A lightweight adapter aligns a vision-language model's (VLM) image encoder with the froze... | [
{
"name": "Guray Ozgur",
"user": "gurayozgur",
"fullname": "Guray Ozgur",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/63a5eeff64f4700278176929/dmIcd1tBGkR33pZZs_CpJ.png",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Mustafa Efe Tamyapar",
"u... | FraunhoferIGD | Fraunhofer IGD | gurayozgur | Guray Ozgur | 0 | 2026-08-21T00:00:00.000Z | true | |||
2401.01173 | 2024-01-03 | En3D: An Enhanced Generative Model for Sculpting 3D Humans from 2D Synthetic Data | [
"Yifang Men",
"Biwen Lei",
"Yuan Yao",
"Miaomiao Cui",
"Zhouhui Lian",
"Xuansong Xie"
] | https://github.com/menyifang/En3D | null | 11 | 9 | 12 | 2024-03-12 | null | null | null | null | 1 | 11 | true | 2024-W01 | 2024-01 | true | 9 | 11 | 34 | 50 | 164 | 219 | We present En3D, an enhanced generative scheme for sculpting high-quality 3D
human avatars. Unlike previous works that rely on scarce 3D datasets or limited
2D collections with imbalanced viewing angles and imprecise pose priors, our
approach aims to develop a zero-shot 3D generative scheme capable of producing
visuall... | [
{
"name": "Yifang Men",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Biwen Lei",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Yuan Yao",
"user": null,
... | null | null | null | akhaliq | AK | null | 2024-01-02T12:06:31.000Z | false | ||
2607.02963 | 2026-07-08 | Parallelized Autoregressive Decoding for Omni-Modal Dense Video Captioning | [
"Wenzheng Zeng",
"Siyi Jiao",
"Chen Gao",
"Hwee Tou Ng",
"Mike Zheng Shou"
] | https://github.com/showlab/PadCaptioner | https://github.com/showlab/PadCaptioner | 27 | 2 | 30 | 2026-07-08 | 16 | 18 | 23 | 27 | 3 | 27 | true | 2026-W28 | 2026-07 | false | 10 | 41 | 37 | 135 | 179 | 601 | Dense video captioning aims to generate temporally grounded descriptions of video events, benefiting both event-level video understanding and generation. In this domain, autoregressive video large language models have emerged as a prevalent paradigm due to their strong generative and cross-modal modeling capacity. Howe... | [
{
"name": "Wenzheng Zeng",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Siyi Jiao",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Chen Gao",
"user": null,
... | NationalUniversityofSingapore | National University of Singapore | wenzhengzeng | Wenzheng Zeng | 33 | 2026-07-03T00:00:00.000Z | false | |||
2411.06559 | 2024-11-21 | Is Your LLM Secretly a World Model of the Internet? Model-Based Planning for Web Agents | [
"Yu Gu",
"Boyuan Zheng",
"Boyu Gou",
"Kai Zhang",
"Cheng Chang",
"Sanjari Srivastava",
"Yanan Xie",
"Peng Qi",
"Huan Sun",
"Yu Su"
] | https://github.com/osu-nlp-group/webdreamer | null | 16 | 2 | 16 | 2024-11-21 | 9 | 10 | 10 | 11 | 0 | 11 | true | 2024-W47 | 2024-11 | false | 6 | 11 | 34 | 58 | 157 | 285 | Language agents have demonstrated promising capabilities in automating
web-based tasks, though their current reactive approaches still underperform
largely compared to humans. While incorporating advanced planning algorithms,
particularly tree search methods, could enhance these agents' performance,
implementing tree s... | [
{
"name": "Yu Gu",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Boyuan Zheng",
"user": "boyuanzheng010",
"fullname": "Boyuan Zheng",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/631a95cfa66151e36e54a9... | null | null | null | akhaliq | AK | 108 | 2024-11-10T18:50:51.000Z | false | ||
2512.02581 | 2025-12-03 | GoRL: An Algorithm-Agnostic Framework for Online Reinforcement Learning with Generative Policies | [
"Chubin Zhang",
"Zhenglin Wan",
"Feng Chen",
"Xingrui Yu",
"Ivor Tsang",
"Bo An"
] | https://github.com/bennidict23/GoRL | null | 4 | 2 | 15 | 2025-12-04 | null | 13 | 13 | 13 | 11 | 13 | true | 2025-W49 | 2025-12 | false | 18 | 48 | 79 | 191 | 277 | 633 | Reinforcement learning (RL) faces a persistent tension: policies that are stable to optimize are often too simple to represent the multimodal action distributions needed for complex control. Gaussian policies provide tractable likelihoods and smooth gradients, but their unimodal form limits expressiveness. Conversely, ... | [
{
"name": "Chubin Zhang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Zhenglin Wan",
"user": "Carlos133386",
"fullname": "Zhenglin Wan",
"avatar": "/avatars/3ef34c3d3db195b90e438a77a9efed37.svg",
"status": "claimed_ver... | NanyangTechnologicalUniversity | Nanyang Technological University | Carlos133386 | Zhenglin Wan | /avatars/3ef34c3d3db195b90e438a77a9efed37.svg | 30 | 2025-12-02T09:49:26.000Z | true | ||
2409.00729 | 2024-09-04 | ContextCite: Attributing Model Generation to Context | [
"Benjamin Cohen-Wang",
"Harshay Shah",
"Kristian Georgiev",
"Aleksander Madry"
] | https://github.com/madrylab/context-cite | null | 14 | 3 | 14 | 2024-09-04 | 9 | 13 | 13 | 13 | 0 | 13 | true | 2024-W36 | 2024-09 | false | 11 | 19 | 30 | 57 | 126 | 254 | How do language models use information provided as context when generating a
response? Can we infer whether a particular generated statement is actually
grounded in the context, a misinterpretation, or fabricated? To help answer
these questions, we introduce the problem of context attribution: pinpointing
the parts of ... | [
{
"name": "Benjamin Cohen-Wang",
"user": "bencw",
"fullname": "Benjamin Cohen-Wang",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/639aaf82a4c528850bba2bfe/nn23r8bsNiOJzVUxAPfo7.png",
"status": "admin_assigned",
"hidden": false
},
{
"name": "Harshay Shah",
"... | null | null | null | akhaliq | AK | 341 | 2024-09-01T14:36:36.000Z | true | ||
2505.16864 | 2025-05-23 | Training-Free Efficient Video Generation via Dynamic Token Carving | [
"Yuechen Zhang",
"Jinbo Xing",
"Bin Xia",
"Shaoteng Liu",
"Bohao Peng",
"Xin Tao",
"Pengfei Wan",
"Eric Lo",
"Jiaya Jia"
] | https://github.com/dvlab-research/Jenga | https://julianjuaner.github.io/projects/jenga/ | 24 | 2 | 24 | 2025-05-23 | 12 | 13 | 20 | 21 | 0 | 21 | true | 2025-W21 | 2025-05 | false | 16 | 47 | 54 | 205 | 196 | 730 | Despite the remarkable generation quality of video Diffusion Transformer
(DiT) models, their practical deployment is severely hindered by extensive
computational requirements. This inefficiency stems from two key challenges:
the quadratic complexity of self-attention with respect to token length and the
multi-step natu... | [
{
"name": "Yuechen Zhang",
"user": "julianjuaner",
"fullname": "zhang yuechen",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/6418554a0956be7233a1023e/9EKN0GoOpcDbvBDmAQEJf.png",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Jinbo Xing",
"user"... | null | null | null | julianjuaner | zhang yuechen | 290 | 2025-05-22T16:21:32.000Z | true | ||
2502.15027 | 2025-02-24 | InterFeedback: Unveiling Interactive Intelligence of Large Multimodal Models via Human Feedback | [
"Henry Hengyuan Zhao",
"Wenqi Pei",
"Yifei Tao",
"Haiyang Mei",
"Mike Zheng Shou"
] | null | null | 7 | 2 | 7 | 2025-02-24 | 5 | 6 | 6 | 6 | 0 | 6 | true | 2025-W09 | 2025-02 | false | 18 | 34 | 84 | 134 | 351 | 502 | Existing benchmarks do not test Large Multimodal Models (LMMs) on their
interactive intelligence with human users which is vital for developing
general-purpose AI assistants. We design InterFeedback, an interactive
framework, which can be applied to any LMM and dataset to assess this ability
autonomously. On top of thi... | [
{
"name": "Henry Hengyuan Zhao",
"user": "hhenryz",
"fullname": "Henry Hengyuan Zhao",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/647d7eb9770c299e56f5b39b/CC5JJgkyLkXOxw-BeT4G5.jpeg",
"status": "admin_assigned",
"hidden": false
},
{
"name": "Wenqi Pei",
"... | null | null | null | akhaliq | AK | null | 2025-02-20T20:27:06.000Z | false | ||
2311.11243 | 2023-11-21 | AutoStory: Generating Diverse Storytelling Images with Minimal Human Effort | [
"Wen Wang",
"Canyu Zhao",
"Hao Chen",
"Zhekai Chen",
"Kecheng Zheng",
"Chunhua Shen"
] | https://github.com/aim-uofa/AutoStory | null | 16 | 3 | 17 | 2024-03-12 | null | null | null | null | 0 | 16 | true | 2023-W47 | 2023-11 | true | 11 | 16 | 32 | 44 | 87 | 174 | Story visualization aims to generate a series of images that match the story
described in texts, and it requires the generated images to satisfy high
quality, alignment with the text description, and consistency in character
identities. Given the complexity of story visualization, existing methods
drastically simplify ... | [
{
"name": "Wen Wang",
"user": "wwen1997",
"fullname": "Wen Wang",
"avatar": "/avatars/04b926a7f2ad091ee00fef0c59903492.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Canyu Zhao",
"user": "Canyu",
"fullname": "Canyu Zhao",
"avatar": "https://cdn-avatars.... | null | null | null | akhaliq | AK | 149 | 2023-11-19T06:07:37.000Z | true | ||
2510.14252 | 2025-10-17 | MoM: Mixtures of Scenario-Aware Document Memories for Retrieval-Augmented Generation Systems | [
"Jihao Zhao",
"Zhiyuan Ji",
"Simin Niu",
"Hanyu Wang",
"Feiyu Xiong",
"Zhiyu Li"
] | https://github.com/MemTensor/MoM | null | 3 | 2 | 3 | 2025-10-17 | 2 | 2 | 2 | 2 | 0 | 2 | true | 2025-W42 | 2025-10 | false | 44 | 51 | 193 | 244 | 756 | 945 | The traditional RAG paradigm, which typically engages in the comprehension of
relevant text chunks in response to received queries, inherently restricts both
the depth of knowledge internalization and reasoning capabilities. To address
this limitation, our research transforms the text processing in RAG from
passive chu... | [
{
"name": "Jihao Zhao",
"user": "Robot2050",
"fullname": "Jihao Zhao",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/658e85bb5b7553ca5c29ba89/KK6UpS9agtrxevvBoup5N.jpeg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Zhiyuan Ji",
"user": null,
... | null | null | null | Robot2050 | Jihao Zhao | 43 | 2025-10-16T03:09:51.000Z | true | ||
2512.22238 | 2025-12-29 | Masking Teacher and Reinforcing Student for Distilling Vision-Language Models | [
"Byung-Kwan Lee",
"Yu-Chiang Frank Wang",
"Ryo Hachiuma"
] | null | null | 30 | 3 | 30 | 2025-12-30 | null | 13 | 16 | 18 | 0 | 18 | true | 2026-W01 | 2025-12 | false | 6 | 15 | 31 | 80 | 211 | 633 | Large-scale vision-language models (VLMs) have recently achieved remarkable multimodal understanding, but their massive size makes them impractical for deployment on mobile or edge devices. This raises the need for compact yet capable VLMs that can efficiently learn from powerful large teachers. However, distilling kno... | [
{
"name": "Byung-Kwan Lee",
"user": "BK-Lee",
"fullname": "Byung-Kwan Lee",
"avatar": "https://cdn-avatars.huggingface.co/v1/production/uploads/657152eb12f162153b50ec9d/qnldHP35PclV0pDz_05q8.jpeg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Yu-Chiang Frank Wang",
... | nvidia | NVIDIA | BK-Lee | Byung-Kwan Lee | null | 2025-12-23T14:40:38.000Z | true | |||
2312.02696 | 2023-12-06 | Analyzing and Improving the Training Dynamics of Diffusion Models | [
"Tero Karras",
"Miika Aittala",
"Jaakko Lehtinen",
"Janne Hellsten",
"Timo Aila",
"Samuli Laine"
] | https://github.com/nvlabs/edm2 | null | 32 | 2 | 34 | 2024-03-12 | null | null | null | null | 1 | 32 | true | 2023-W49 | 2023-12 | true | 2 | 27 | 10 | 99 | 30 | 293 | Diffusion models currently dominate the field of data-driven image synthesis
with their unparalleled scaling to large datasets. In this paper, we identify
and rectify several causes for uneven and ineffective training in the popular
ADM diffusion model architecture, without altering its high-level structure.
Observing ... | [
{
"name": "Tero Karras",
"user": "tkarras",
"fullname": "Tero Karras",
"avatar": "/avatars/cb95f524dd1c80aa6ec8af977c5f2606.svg",
"status": "extracted_confirmed",
"hidden": false
},
{
"name": "Miika Aittala",
"user": "miika",
"fullname": "Miika Aittala",
"avatar": "/avata... | null | null | null | akhaliq | AK | 853 | 2023-12-05T11:55:47.000Z | false | ||
2503.22673 | 2025-04-01 | ActionStudio: A Lightweight Framework for Data and Training of Large Action Models | [
"Jianguo Zhang",
"Thai Hoang",
"Ming Zhu",
"Zuxin Liu",
"Shiyu Wang",
"Tulika Awalgaonkar",
"Akshara Prabhakar",
"Haolin Chen",
"Weiran Yao",
"Zhiwei Liu",
"Juntao Tan",
"Juan Carlos Niebles",
"Shelby Heinecke",
"Huan Wang",
"Silvio Savarese",
"Caiming Xiong"
] | https://github.com/SalesforceAIResearch/xLAM | null | 12 | 2 | 13 | 2025-04-01 | 8 | 9 | 11 | 12 | 1 | 12 | true | 2025-W14 | 2025-04 | false | 14 | 27 | 75 | 124 | 246 | 456 | Action models are essential for enabling autonomous agents to perform complex
tasks. However, training large action models remains challenging due to the
diversity of agent environments and the complexity of agentic data. Despite
growing interest, existing infrastructure provides limited support for
scalable, agent-spe... | [
{
"name": "Jianguo Zhang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Thai Hoang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Ming Zhu",
"user": null,... | null | null | null | jianguozhang | Jianguo Zhang | null | 2025-03-28T17:58:33.000Z | false | ||
2603.17375 | 2026-03-19 | Stereo World Model: Camera-Guided Stereo Video Generation | [
"Yang-Tian Sun",
"Zehuan Huang",
"Yifan Niu",
"Lin Ma",
"Yan-Pei Cao",
"Yuewen Ma",
"Xiaojuan Qi"
] | https://github.com/SunYangtian/StereoWorld | https://sunyangtian.github.io/StereoWorld-web/ | 11 | 2 | 11 | 2026-03-19 | 8 | 9 | 10 | 11 | 0 | 11 | true | 2026-W12 | 2026-03 | false | 18 | 33 | 92 | 194 | 324 | 745 | We present StereoWorld, a camera-conditioned stereo world model that jointly learns appearance and binocular geometry for end-to-end stereo video generation.Unlike monocular RGB or RGBD approaches, StereoWorld operates exclusively within the RGB modality, while simultaneously grounding geometry directly from disparity.... | [
{
"name": "Yang-Tian Sun",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Zehuan Huang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Yifan Niu",
"user": nu... | null | null | null | huanngzh | zehuan-huang | 101 | 2026-03-18T05:42:22.000Z | false | ||
2607.22529 | 2026-07-27 | Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills | [
"Siyuan Huang",
"Pengyu Cheng",
"Haotian Liu",
"Tao Chen",
"Yihao Liu",
"Jingwei Ni",
"Shijie Zhou",
"Ziyi Yang",
"Gangwei Jiang",
"Mengyu Zhou",
"Yu Cheng",
"Xiaoxi Jiang",
"Guanjun Jiang"
] | https://github.com/Qwen-Applications/skill-self-play | null | 48 | 1 | 48 | 2026-07-27 | 30 | 35 | 46 | 47 | 1 | 47 | true | 2026-W31 | 2026-07 | false | 2 | 17 | 27 | 138 | 100 | 601 | LLM training is shifting from manual design and annotation to interaction-driven self-evolution. However, existing self-evolutionary methods face a fundamental dilemma between task diversity and verification reliability: environment-bound methods obtain precise feedback but confine learning to narrow domains, while ope... | [
{
"name": "Siyuan Huang",
"user": "chamber111",
"fullname": "Siyuan Huang",
"avatar": "/avatars/92918bf8913012a3f005f09e03b381c2.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Pengyu Cheng",
"user": "Linear95",
"fullname": "Pengyu Cheng",
"avatar": "/av... | QwenBusinessUnit | Qwen Business Unit | taesiri | taesiri | 149 | 2026-07-24T00:00:00.000Z | false | |||
2602.13191 | 2026-02-16 | CoPE-VideoLM: Codec Primitives For Efficient Video Language Models | [
"Sayan Deb Sarkar",
"Rémi Pautrat",
"Ondrej Miksik",
"Marc Pollefeys",
"Iro Armeni",
"Mahdi Rad",
"Mihai Dusmanu"
] | null | https://sayands.github.io/cope/ | 34 | 2 | 34 | 2026-02-16 | 20 | 26 | 29 | 29 | 0 | 29 | true | 2026-W08 | 2026-02 | false | 6 | 32 | 18 | 137 | 138 | 785 | Video Language Models (VideoLMs) empower AI systems to understand temporal dynamics in videos. To fit to the maximum context window constraint, current methods use keyframe sampling which can miss both macro-level events and micro-level details due to the sparse temporal coverage. Furthermore, processing full images an... | [
{
"name": "Sayan Deb Sarkar",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Rémi Pautrat",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Ondrej Miksik",
"us... | microsoft | Microsoft | sayandsarkar | Sayan Deb Sarkar | /avatars/c26c03fa920d857120f03c9ccb9f1d7a.svg | null | 2026-02-13T18:57:31.000Z | false | ||
2608.18027 | 2026-08-21 | Chain-of-Experience for Continual LLM Improvement | [
"Haoqin Tu",
"Yunhao Fang",
"Yizhong Wang",
"Cihang Xie",
"Shen Yan"
] | null | null | 11 | 2 | 11 | 2026-08-21 | 7 | 7 | 8 | 9 | 0 | 9 | true | 2026-W34 | 2026-08 | false | 18 | 25 | 93 | 153 | 417 | 647 | Humans continuously learn from experience, whereas conventional large language model (LLM) evaluations ignore the models' ability to improve through inference-time interaction. In this paper, we study how LLMs learn from iterative experience at test time, a setting we refer to as Chain-of-Experience (CoE), where models... | [
{
"name": "Haoqin Tu",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Yunhao Fang",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Yizhong Wang",
"user": null... | UCSC-VLAA | UCSC-VLAA | PahaII | Haoqin Tu | null | 2026-08-18T00:00:00.000Z | false | |||
2608.30428 | 2026-09-01 | Lies We Can See: Joint Verbal and Non-Verbal Deception by VLM Agents in Embodied Social Interactions | [
"Jaewoo Ahn",
"Junseo Kim",
"Hyunseo Kim",
"Heeseung Yun",
"Jaehyeon Son",
"Zsolt Kira",
"Gunhee Kim"
] | https://github.com/JunseoKim0103/Lies-We-Can-See | https://junseokim0103.github.io/Lies-We-Can-See/ | 13 | 2 | 14 | 2026-09-07 | null | null | null | 14 | 1 | 14 | true | 2026-W36 | 2026-09 | false | 16 | 39 | 93 | 166 | 448 | 786 | Strategic deception by LLM and VLM agents has emerged as a central AI alignment and safety concern. Social-deduction games (where each player holds a hidden role and communicates with others to deduce identities) serve as the canonical testbed, particularly in multi-agent settings. Existing testbeds, however, are text-... | [
{
"name": "Jaewoo Ahn",
"user": "ahnpersie",
"fullname": "Jaewoo Ahn",
"avatar": "/avatars/038894bc72b92ec3f4ecb096cc60b60a.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Junseo Kim",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
... | SeoulNatlUniv | Seoul National University | ahnpersie | Jaewoo Ahn | /avatars/038894bc72b92ec3f4ecb096cc60b60a.svg | 8 | 2026-08-31T00:00:00.000Z | true | ||
2402.04744 | 2024-02-08 | Progressive Gradient Flow for Robust N:M Sparsity Training in Transformers | [
"Abhimanyu Rajeshkumar Bambhaniya",
"Amir Yazdanbakhsh",
"Suvinay Subramanian",
"Sheng-Chun Kao",
"Shivani Agrawal",
"Utku Evci",
"Tushar Krishna"
] | https://github.com/abhibambhaniya/progressive_gradient_flow_nm_sparsity | null | 2 | 1 | 2 | 2024-03-12 | null | null | null | null | 0 | 2 | true | 2024-W06 | 2024-02 | true | 14 | 14 | 65 | 65 | 256 | 256 | N:M Structured sparsity has garnered significant interest as a result of
relatively modest overhead and improved efficiency. Additionally, this form of
sparsity holds considerable appeal for reducing the memory footprint owing to
their modest representation overhead. There have been efforts to develop
training recipes ... | [
{
"name": "Abhimanyu Rajeshkumar Bambhaniya",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hidden": false
},
{
"name": "Amir Yazdanbakhsh",
"user": "ayazdan",
"fullname": "Amir Yazdanbakhsh",
"avatar": "/avatars/76e5e7b51515b58e207c7d2a4b360d30.svg",
... | null | null | null | akhaliq | AK | 11 | 2024-02-07T10:55:59.000Z | false | ||
2403.01807 | 2024-03-05 | ViewDiff: 3D-Consistent Image Generation with Text-to-Image Models | [
"Lukas Höllein",
"Aljaž Božič",
"Norman Müller",
"David Novotny",
"Hung-Yu Tseng",
"Christian Richardt",
"Michael Zollhöfer",
"Matthias Nießner"
] | https://github.com/facebookresearch/viewdiff | null | 8 | 1 | 9 | 2024-03-12 | null | null | null | 6 | 1 | 6 | true | 2024-W10 | 2024-03 | false | 8 | 11 | 41 | 48 | 190 | 218 | 3D asset generation is getting massive amounts of attention, inspired by the
recent success of text-guided 2D content creation. Existing text-to-3D methods
use pretrained text-to-image diffusion models in an optimization problem or
fine-tune them on synthetic data, which often results in non-photorealistic 3D
objects w... | [
{
"name": "Lukas Höllein",
"user": "lukasHoel",
"fullname": "Lukas Hoellein",
"avatar": "/avatars/162a798948bfe3df8e788aba5d024ffb.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Aljaž Božič",
"user": null,
"fullname": null,
"avatar": null,
"status":... | null | null | null | akhaliq | AK | 382 | 2024-03-04T07:57:05.000Z | false | ||
2410.11842 | 2024-10-18 | MoH: Multi-Head Attention as Mixture-of-Head Attention | [
"Peng Jin",
"Bo Zhu",
"Li Yuan",
"Shuicheng Yan"
] | https://github.com/skyworkai/moh | null | 20 | 2 | 22 | 2024-10-18 | 8 | 16 | 19 | 20 | 2 | 20 | true | 2024-W42 | 2024-10 | false | 11 | 34 | 34 | 117 | 116 | 482 | In this work, we upgrade the multi-head attention mechanism, the core of the
Transformer model, to improve efficiency while maintaining or surpassing the
previous accuracy level. We show that multi-head attention can be expressed in
the summation form. Drawing on the insight that not all attention heads hold
equal sign... | [
{
"name": "Peng Jin",
"user": "Chat-UniVi",
"fullname": "Peng Jin",
"avatar": "/avatars/579e468334102472d870875fe40302e6.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Bo Zhu",
"user": null,
"fullname": null,
"avatar": null,
"status": null,
"hid... | null | null | null | Chat-UniVi | Peng Jin | /avatars/579e468334102472d870875fe40302e6.svg | 310 | 2024-10-15T17:59:44.000Z | true | |
2306.01160 | 2023-06-05 | Faster Causal Attention Over Large Sequences Through Sparse Flash Attention | [
"Matteo Pagliardini",
"Daniele Paliotta",
"Martin Jaggi",
"François Fleuret"
] | https://github.com/epfml/dynamic-sparse-flash-attention | null | 1 | 2 | 1 | 2024-03-12 | null | null | null | null | 0 | 1 | true | 2023-W23 | 2023-06 | true | 8 | 10 | 61 | 69 | 239 | 257 | Transformer-based language models have found many diverse applications
requiring them to process sequences of increasing length. For these
applications, the causal self-attention -- which is the only component scaling
quadratically w.r.t. the sequence length -- becomes a central concern. While
many works have proposed ... | [
{
"name": "Matteo Pagliardini",
"user": "MatPag",
"fullname": "Matteo Pagliardini",
"avatar": "/avatars/38a6c66bce327eded16993932b62986b.svg",
"status": "claimed_verified",
"hidden": false
},
{
"name": "Daniele Paliotta",
"user": "dpstart",
"fullname": "Daniele Paliotta",
... | null | null | null | akhaliq | AK | 152 | 2023-06-01T21:33:59.000Z | true |
End of preview. Expand in Data Studio
Paper Pulse data
The day-by-day upvote history of every Hugging Face Daily Papers entry since 2024-03-12, updated every day. It powers Paper Pulse (Space).
Built from the past revisions of hysts-bot-data/daily-papers-stats and the paper list in hysts-bot-data/daily-papers, both by @hysts, plus paper details from the public Daily Papers API. Thanks to hysts for keeping that record and for releasing it under CC0 1.0.
Files
| File | Rows | What it holds |
|---|---|---|
series.parquet |
one per paper per day | id (arXiv id), day, upvotes, comments: the count shown on the paper's page that day (last snapshot of the day, UTC) |
papers.parquet |
one per paper | title, Daily Papers date, authors (with Hugging Face accounts when linked), organization, submitter, GitHub repo and stars, abstract (summary), upvotes on day 0/1/3/7, up_close (7 days after the Daily Papers date), peak, purged, and rank within its day, ISO week and month |
purges.parquet |
one per event | days when the Hub removed votes from many papers at once: papers affected (n_drop), votes removed, and whether it was deep |
hourly.parquet |
one per paper per hour | id, ts (UTC), upvotes for the papers featured in the last 10 days, from the hourly snapshots of daily-papers-stats |
age_stats.json |
how papers usually do by age: the distribution of upvotes at the end of day 0..7 after their Daily Papers date, and what papers with a given count at day k ended their first week with (ratios by count bin) | |
meta.json |
last day, number of papers, purge days |
Notes
- Upvotes can go down. People unvote, and on a few days the Hub removed votes from many papers at once (2025-09-12, 2025-12-23, 2026-09-16, 2026-09-18; the cause was not announced). The series keeps the raw value of each day and never corrects it. A day counts as a purge when at least max(250, 2% of papers) dropped.
- Rankings use
up_close, the upvotes 7 days after the Daily Papers date, so a later purge does not rewrite past rankings. Papers younger than 7 days are ranked by their current count, and so are the papers whose Daily Papers date is before the history starts (March 12, 2024), which are flagged withearly. - The source has two collection gaps (13 and 7 days). Days without a snapshot have no row.
- Only papers that were featured on Daily Papers are included (about 18,500), not every paper on the Hub.
License
The series, rankings and purge records are released under CC BY 4.0. The source datasets by @hysts are CC0 1.0. Paper titles and abstracts (title, summary) belong to their authors and are included only to identify each paper; they are not covered by this dataset's license.
- Downloads last month
- 264