paper_id stringlengths 10 10 | items listlengths 4 82 | gt float64 1 9.5 |
|---|---|---|
gbJNFxcicC | [
"strength: Easy to read paper, tackling a relevant societal problem.",
"strength: The dataset gathered and annotated is quite large for a medical dataset, which is excellent work and sure to be useful to those who can make use of it in the future. Although I assume it will not be possible to make the dataset open... | 1.666667 |
B8akWa62Da | [
"strength: Originality:The HPFL framework proposed in this paper innovatively introduces a plug-in selection mechanism into federated learning, realizes the bridge between general models and personalized models, and solves the performance balance problem of traditional GFL and PFL. This is a new attempt in federate... | 6 |
e9bEoxNiTJ | [
"strength: The content is well-organized and easy to follow.",
"strength: The motivation is well-established and the effectiveness of their solution is verified by extensive experiment.",
"strength: The proposed architecture achieved competitive performance on a wide range of tasks, while maintaining competitiv... | 5.333333 |
DCg9r2DKKe | [
"strength: The approach seems reasonable to me, as various existing works introduce STL robustness loss to enforce formal properties in training modules.",
"strength: The paper raises an interesting problem about ensuring safe learning from expert demonstrations, in an imitation learning framework. This is a sign... | 2.5 |
VIUisLx8lQ | [
"strength: The approach to incorporate multiple reasoning types (deductive, inductive, abductive, and analogical) into consideration for a meta-thinker for LLM reasoning is novel and seems like a generalizable approach for most LLMs.",
"strength: There were clear quantitative gains for this approach.",
"strengt... | 6 |
gU58d5QeGv | [
"strength: This study tackles an important topic of reducing the computational cost of text-to-image diffusion models.",
"strength: The method introduced in the study is both innovative and efficient, offering clear results and validating its effectiveness through extensive evaluations.",
"strength: The paper i... | 8 |
wGqf7YMF8R | [
"strength: strong performance improvement compared with direct CoT thinking",
"strength: The paper presents a new approach that facilitates deliberate, slow reasoning. (Compared to previous methods like CoT/PAL, ) this method automatically breaks down complex problems into smaller sub-tasks and designs a dynamic ... | 4.75 |
itGkF993gz | [
"strength: The figures are nice and informative.",
"strength: Various PPI baselines are considered, including both sequence- and structure-based approaches.",
"strength: The author describes the fundamental algorithm well; and they seem to give all relevant information to understand and reproduce their algorith... | 5.666667 |
84WmbzikPP | [
"strength: As far as I can tell, the approach used in the paper of embedding the molecular structure elucidation task in the Stiefel manifold is original. This approach leverages manifold geometry to respect exact moment constraints, improving over traditional Euclidean diffusion models.",
"strength: In general, ... | 7 |
RAdBtquPiI | [
"strength: The authors study how to ensure safety constraint satisfaction during training and deployment by assuming a constrained world model. I recongize this is a new constrained learning setting.",
"strength: The authors apply the Bender's oracle optimization to constrained MDP problems by enforcing hard cons... | 3.4 |
B37UmlxsaP | [
"strength: The paper includes extensive experiments with several ablations that support their hypothesis.",
"strength: The use of outlier detection in summarization models may help improve their performance, although this was not explored in the paper.",
"strength: The paper present a clear formalization of the... | 2.5 |
bESxQeXTlo | [
"strength: A simple pipeline that works reasonably well. Each component of the overall pipeline is simple and easy to understand",
"strength: Compared to structural anomaly detection, logical anomaly detection remains relatively under-explored. This paper proposes a simple and easy-to-implement solution.",
"str... | 3 |
J6nKxekCCo | [
"strength: The study covers five LLMs and 50 intersectional identities.",
"strength: The authors guard against robustness issues when prompting LLMs for bias by also considering diverse rephrasings of their prompts [1].",
"strength: The authors apply the methodology of [2] to consider disparities in the warmth ... | 3 |
hCfhfwSfCg | [
"strength: It is an interesting paper which combines large language model (LLM) guidance with model-based reinforcement learning (RL) to tackle sparse-reward environments.",
"strength: This approach provides a novel perspective by using LLMs to guide exploration goals, addressing sparse reward in RL.",
"strengt... | 2 |
6s5uXNWGIh | [
"strength: Originality. To my knowledge, this benchmark is the first that is designed to evaluate ML engineering capabilities of AI agents.",
"strength: Quality. The paper has done a good-faith effort to benchmark using relevant models and mitigate potential issues (e.g., contamination and plagiarism). I liked th... | 8 |
YuYxoaL7YX | [
"strength: The paper generalizes the arrival dynamics of the inventory system and provides simulation results to support their method.",
"strength: The paper considers a problem motivated from real inventory control problems. The formulation of the problem is sound.",
"strength: Discusses an important but over... | 3 |
zd5Knrtja4 | [
"strength: Paper writing is good. There is no difficulty in understanding the paper.",
"strength: The proposed method is straightforward and achieves state-of-the-art performance.",
"strength: Comprehensive ablation studies and other evaluations are conducted to check the effects of the model",
"strength: The... | 5 |
UyBMzsFThf | [
"strength: The proposed method is simple and can be easily extended as a main objective during CLIP pretraining (not just as a finetuning objective).",
"strength: The proposed method shows significant improvement on tasks involving difference based classification.",
"strength: The paper proposes an effective me... | 5.25 |
cSnbM9SIJJ | [
"strength: The system architecture is well-designed with a clear separation of components and modular design.",
"strength: The scaling-up simulation size problem is an important challenge in multi-agent systems research.",
"strength: Extensive experimental analysis across different LLMs, prompts, and agent conf... | 3 |
2uPZ4aX1VV | [
"strength: The introduction of the *Null counterfactual interaction assumption* could be an important contribution, improving sample efficiency across various domains, particularly in manipulation tasks where interaction is minimal.",
"strength: The method details engineering practices to make the approach both m... | 6.75 |
3IFRygQKGL | [
"strength: The need for efficiency in decision making in RL is clear, as single-step actions are slow and computationally expensive (even more so in slow simulators). Thus, the problem addressed by OptionZero is clear and its existence is well-motivated.",
"strength: Additionally, since much prior work in options... | 7.5 |
fvTaoyH96Z | [
"strength: The paper is written clearly and well organized.",
"strength: The motivation for the problem of environmental generalization is intuitive and interesting.",
"strength: The framework provides a nice formalism for environment generalization.",
"strength: The framework proposed in this paper for descr... | 2.333333 |
EWQaqDgXgr | [
"strength: The paper tackles a timely and practically-relevant problem supported by a fair amount of experiments. Shielded generation of text-to-image diffusion models is an area with limited prior research, making this work particularly valuable.",
"strength: Overall, the paper is clearly written and easy to fol... | 5.75 |
gbrHZq07mq | [
"strength: The paper contribute to a fun connection between formal language theory in order to analyze the expressivity of transformers. This line of research sounds more like research performed in TCS than in IA tracks but since it is apply to IA-defined model it makes some sense. Understanding the expressivity of... | 5.6 |
DZqic2sPTY | [
"strength: The paper pioneers the integration of TDA into the study of temporal graphs, which represents a significant methodological advancement, potentially unlocking new insights into graph structure and evolution over time.",
"strength: The paper is commended for its clear presentation style and the provision... | 7 |
hWlCc7Iksi | [
"strength: The proposed method is aware of the advantages of using spatiotemporal data and randomness for pretraining, which yields better results on Kinetics dataset.",
"strength: The proposed method exhibits an improved training time and memory usage compared to VideoMAE, while achieving the same level of perfo... | 3.4 |
ghyeMoj1gK | [
"strength: The idea to have clients control the update process is interesting and novel",
"strength: The solution is sensible and the results are promising",
"strength: The philosophical motivation presented by the manuscript is quite compelling: a client-centric federated framework.",
"strength: Handling loc... | 5 |
d94x0gWTUX | [
"strength: The paper addresses an important issue in reward modeling by introducing a tool-augmented approach to enhance the effectiveness of RMs.",
"strength: The proposed methodology of integrating external tools into RMs is innovative and practical, allowing for dynamic decision-making and reasoning processes.... | 7.333333 |
PstM8YfhvI | [
"strength: MorphoDiff method shows novel methods combining latent diffusion models with perturbation embedding to predict cellular morphologies, generalizing across cell types, drugs, and genetic alterations better than the standard stable diffusion model, and demonstrating dual capability with gene knock-in/out an... | 6.75 |
KnoS9XxIlK | [
"strength: Improving the predictability of LLMs at different scales has been a very important task for many years. This paper attempts to make the projections very granular, in particular, authors propose an empirically tuned power law that allows to predict the training loss for a given model at every individual s... | 6 |
9DvDRTTdlu | [
"strength: The paper is in general well organized and easy to follow.",
"strength: The DDS distillation loss with the binary mask is shown to be effective.",
"strength: The paper provides a well-motivated solution to the problem of editing pre-trained 3D implicit networks. The proposed approach effectively pres... | 5.5 |
SW6IIpPr9I | [
"strength: This manuscript models label errors as stochastic variables and derives embedding space constraints to arrange instances based on their probabilistically associated ranks.",
"strength: This manuscript proposes outlier detection and relabeling schemes to pinpoint and mitigate instances with extreme labe... | 4.25 |
gqtbL7j2JW | [
"strength: The problem setting appears novel and useful. Especially, when there are hundreds of models in a model zoo and one needs to find what model could be used to produce a given image.",
"strength: The method is straightforward, the exposition is very clear, and easy to read.",
"strength: Experiments show... | 5 |
a8XwgTZzE0 | [
"strength: The topic of delayed generalization is interesting and suitable to the ICLR audience.",
"strength: Viewing gradient descent updates to neural network parameters as the evolution of a dynamical system is an interesting (but not novel) perspective.",
"strength: Trying to relate generalisation phenomena... | 2 |
E4LAVLXAHW | [
"strength: A huge number of watermarking papers have come out recently.",
"strength: Many of them ask whether their watermarks harm generation quality by performing experimental evaluations, but these are inherently limited: There is no way to experimentally guarantee that the watermark will preserve the quality ... | 7 |
xayT1nn8Mg | [
"strength: Incorporates appropriate symmetries for (temporal) 3D point cloud learning. In particular, the invariance to time reparameterization is key because the underlying MD simulations might be prone to random restarts and miscellaneous artefacts preventing them from being a smooth \"video\" (ie, the MD itself ... | 6.666667 |
MMHqnUOnl0 | [
"strength: The authors combine biological insight with language modeling.",
"strength: The proposed HELM is solid and performs well for mRNA-related tasks. It achieves SOTA performance on several practical and impactful tasks, e.g., mRNA sequence design, and sequence region annotation.",
"strength: The authors ... | 5.75 |
Nx4PMtJ1ER | [
"strength: Provides a solid framework for causal inference in continuous-time SDEs, going beyond the limitations of traditional discrete-time models.",
"strength: Introduces a practical CI test using signature kernels, suited for handling path-dependent random variables.",
"strength: Shows strong performance wi... | 8 |
qbSoiHLEK0 | [
"strength: LLM2Features represents an innovative use of LLMs for automated feature engineering, providing a unique contribution to AutoML by reducing reliance on human-supplied domain knowledge.",
"strength: By using interpretable feature generation, the method allows users to understand the logical basis of gene... | 3 |
emMMa4q0qw | [
"strength: Very important problem: The paper broadens our understanding of the ventral stream by exploring if spatial tasks, not just object categorization, can also drive effective visual cortex models. This is impactful for both building better models of the cortex, and also provides a computational feasibility a... | 7 |
AfnsTnYphT | [
"strength: While the results agree with empirical intuition that inductive biases in the model reflecting properties of the data should improve sample complexity, it is interesting to see such a result made crisp for vision models, and with specific focus on the locality and translation invariance properties of vis... | 7.5 |
t9lS1lX9FQ | [
"strength: very well written paper about a pressing topic (graph quantization)",
"strength: extensive results help provide details about the method",
"strength: I'm confident the method seems to work well (but perhaps doesn't actually win on baselines, see weaknesses)",
"strength: The paper introduces a new f... | 6 |
9pW2J49flQ | [
"strength: The paper presents an interesting approach to learn policies to satisfy omega-regular specifications based on visiting accept states in an automaton without discounting states between the visits.",
"strength: It incorporates policies parameterized as neural networks.",
"strength: It uses the structur... | 8 |
7lpDn2MhM2 | [
"strength: All code and datasets are public.",
"strength: Well written.",
"strength: Is very clear about how baselines should be compared (not everything can be directly compared, but they still include them).",
"strength: Introduces all benchmarks and evaluation tasks/metrics in the paper with a quick overvi... | 6.333333 |
HAD6iZxKuh | [
"strength: WMAdapter is compatible with various diffusion models and watermarking decoders, enhancing its adaptability for real-world applications.",
"strength: The hybrid fine-tuning scheme is innovative, effectively reducing artifacts.",
"strength: The paper is generally well-written and easy to follow.",
"... | 5.2 |
eY5JNJE56i | [
"strength: The paper is well written and the structure is clear. The idea is supported by both theoretical justification and relatively sufficient empirical study.",
"strength: The authors made a great effort in the rebuttal process and addressed all my concerns.",
"strength: First, I find that the problem tack... | 6.75 |
GQgPj1H4pO | [
"strength: Enabling VidSGG training using only video captions.",
"strength: Reduction in annotation costs compared to existing approaches.",
"strength: Well-identified key challenges in adapting image-based weak supervision to videos.",
"strength: Modular architecture with TCS and ADV components.",
"strengt... | 6 |
1Uem0nAWK0 | [
"strength: A novel inference-time model editing technique using Alignment Vectors that allows dynamic adjustment of LLM outputs along preference dimensions without retraining or complex prompt engineering",
"strength: A substantial synthetic dataset (38k examples) spanning three domains and three proficiency leve... | 4.25 |
HFAIxjBB6K | [
"strength: The motivation and the proposed method is easy to follow.",
"strength: The proposed criteria for gradient suppressed softmax is general.",
"strength: This paper offers a new perspective, namely the gradient optimization of Softmax, to discuss the issue of impaired diversity in autoregressive generate... | 4.75 |
BhxsjonZ0z | [
"strength: The research direction of considering classical outlier detection algorithms is less exploited.",
"strength: The idea of algorithm decomposition and NN approximation for each component is interesting.",
"strength: The challenges and approaches for designing the FL version of non-NN OD algorithms are ... | 3.333333 |
LDu822E45Q | [
"strength: The research question of identifying the right benchmark is intriguing, practical, and crucial for both the research community and industry, as it enables a fair, efficient, and cost-effective evaluation protocol for model development.",
"strength: The problem studied---evaluation accuracy and efficien... | 4.25 |
nTZOIlf8YH | [
"strength: They consider some loss functions to capture the discrepancies between predicted and true decision problems.",
"strength: They explore a particular case of web advertisement allocation within the Anonymous App, aiming to optimize overall click metrics and enhance user visitation on the following day.",... | 2.333333 |
UVnD9Ze6mF | [
"strength: Alignment with actual regulations: A novel feature of this benchmark is its grounding in real-world regulations. By basing AIR-Bench on key regulations from the EU, US, and China, it addresses critiques that existing AI safety benchmarks lack practical relevance, providing a benchmark that better reflect... | 7.5 |
Tzh6xAJSll | [
"strength: Admirably formalizes the scaling laws in Transformers as a memorization/memory retrieval task in Associative Memories",
"strength: The paper clearly and thoroughly defines a \"sandbox\" problem setting where we can study scaling laws (of discrete data domains, like the vocabulary tokens in NLP) using p... | 7.6 |
CGbfokGFP7 | [
"strength: It extends the 3D representation from M-SDF to support color and PBR materials and integrates it into the generative framework, enabling the generation of 3D assets containing PBR materials from text and image conditions. This introduces a new approach to 3D representation for native 3D generative tasks.... | 4.75 |
cqTUJRlcLU | [
"strength: Novelty: Authors are first to study it for transformer models.",
"strength: Contributions seems to be rigorously proved.",
"strength: First theoretical analysis of benign overfitting in attention mechanisms and unique analysis of token selection behavior and how attention handles noisy labels. Severa... | 5.8 |
6GATHdOi1x | [
"strength: The intuition of incorporating ranking loss into the DM recommender makes sense. The development of the proposed method looks reasonable and (to my knowledge) correct.",
"strength: The paper is clearly presented and easy to follow.",
"strength: Experiments and ablation studies are mostly solid.",
"... | 5.75 |
lXE5lB6ppV | [
"strength: This paper utilizes a range of task-specific datasets to explore the posed questions and provide sufficient evaluations.",
"strength: This paper shows a new finding from the task-specific fine-tuning of aligned LLMs.",
"strength: This paper proposes a corresponding to address the threats when fine-tu... | 5.75 |
xIHi5nxu9P | [
"strength: The paper is reasonably clear and proposes simple yet interesting idea which appears to work well on selected synthetic/small scale experiments.",
"strength: The authors provide the code for the experiments (which I did not reviewed).",
"strength: The figure in page 1 nicely summarises the benefit of... | 7.2 |
zuOnOAHBMy | [
"strength: HELENE’s integration of annealed A-GNB gradients and layer-wise clipping shows the authors' awareness of the specific computational nuances of finetuning LLM architectures, the discussion of EMA showed the and other related discussion showed good motivation.",
"strength: Plus, memory saving is always c... | 4 |
OVu9DsOjgH | [
"strength: The paper presents an interesting and potentially impactful approach to OOD generalization in time series classification using frequency decomposition.",
"strength: The frequency-based separation into time-deterministic and time-stochastic components is theoretically motivated and could offer a meaning... | 5.2 |
tUiYbVqcuQ | [
"strength: (Minor Strength - Novelty - Critic Aggregation) To the best of my limited knowledge on the subject, A2FC provides a novel approach for MARL. I specifically have not seen a method that aggregates individual critics by averaging out their parameters like in A2FC.",
"strength: The authors seem to have a g... | 3 |
IMWYNVBHob | [
"strength: The problem studied is interesting and novel. There is no existing work on explaining heterogeneous graph neural networks.",
"strength: Experimental results demonstrate the effectiveness of the proposed method",
"strength: The paper has a theoretical analysis on the noise amplification effect in hete... | 3 |
FK8tl47xpP | [
"strength: The theoretical part of the paper is well-written and provides a very clear story. The theoretical framework also provides a good starting point for further generalising analysis of L2O schemes, especially when moving to non-convex settings.",
"strength: The proposed preconditioner learning is memory-e... | 6.25 |
WxLwXyBJLw | [
"strength: The single step sampling results in a greatly improved speed without significantly affecting the model performance.",
"strength: Adresses the problem of long inference time by bypassing the inference ODE integration.",
"strength: The most important benefit of the proposed sampling approach is its spe... | 3.25 |
krx55l2A6G | [
"strength: D-SNR is a useful metric to detect disaggregation by malicious server",
"strength: SEER is an attack that can efficiently extract information and bypass D-SNR detection.",
"strength: The paper is in general well-written and self-contained and provides a good summarization of prior studies in this fie... | 6.2 |
FDsWd0NOB5 | [
"strength: The authors tackled an interesting problem that has not been extensively studied.",
"strength: The writing is clear and easy to follow.",
"strength: The authors provided both qualitative and some quantitative metrics in the evaluation of their framework.",
"strength: Strategy: The strategy of using... | 5.5 |
EpnZEzYDUT | [
"strength: The diffusion approach makes sense. Using a diffusion model to enrich offline datasets could support several other MARL algorithms.",
"strength: The employed techniques align with recent advancements in diffusion-based RL.",
"strength: The paper is well-written and easy to follow.",
"strength: The ... | 6.5 |
8HQS1X2AK4 | [
"strength: The use of ensemble reweighting for test-time alignment is a novel solution to the underspecified task problem, offering quick adaptation without additional training.",
"strength: HYRE leverages scalable ensemble architectures, making it feasible to apply this approach to large-scale, pretrained models... | 5.333333 |
l5HEECYJ3i | [
"strength: The paper provides extensive empirical evidence demonstrating the effectiveness of the proposed method over considered baseline methods.",
"strength: The overall approach is well-motivated and clearly stated.",
"strength: Empirical evaluations look good.",
"strength: This paper is well-structured. ... | 3.5 |
vgt2rSf6al | [
"strength: The authors employ a generative fMRI encoding model to synthesize individual fMRI signals corresponding to concept-oriented visual stimuli, addressing the inherent one-to-many correspondence issue between visual stimuli and fMRI recordings.",
"strength: The paper is well-structured, with a clear formul... | 5.75 |
oRfHv642qD | [
"strength: The paper adds to the literature of biologically-inspired inference. The extension introduced is simple and elegant and allows generalizing P-VAE to dynamically trade-off performance for compute.",
"strength: The experiments show that despite IP-VAE being order of magnitude more parameter efficient, th... | 4.4 |
2J18i8T0oI | [
"strength: The identification of the \"Off-by-One motif\" in Mamba models is a unique contribution that highlights nuanced differences between architectures.",
"strength: The introduction of a complexity-based interpretation for understanding feature similarity differences is innovative.",
"strength: The circui... | 6.5 |
awvJBtB2op | [
"strength: Novelty. Unlike many previous works that focus exclusively on rigid or soft bodies, this work combines them into a single framework. The manuscript provides an extensive related works section for situating this work within the field, which was helpful.",
"strength: Clarity. The manuscript was a pleasur... | 7.5 |
XWfjugkXzN | [
"strength: Partial observability is a practical issue in real-world decision making problems, and this paper tackles this issue from a perspective based on information set.",
"strength: The paper addresses a problem of importance to the game theory community, namely, value estimation of information sets in games.... | 1.666667 |
AEi2wyAMyb | [
"strength: The proposed method outperforms many other SSL algorithms on common SSL benchmark datasets.",
"strength: Ablation studies shows that pseudo-labels benefit from ReLU projections instead of softmax which is an obvious choice for projection onto the probability simplex. It's good that the paper experiment... | 5.333333 |
1CRu6bGx25 | [
"strength: The paper presents interesting and novel applications of saliency maps to model quantization and model merging.",
"strength: The paper strongly motivates the usage of saliency maps.",
"strength: FI offers a theoretically grounded approach to assessing parameter and input sensitivity, which can suppor... | 3.666667 |
5XUlfPcQnG | [
"strength: The area of work addressed by the paper is very much needed - HVAC is one of the largest contributors of emissions in buildings, but simulators that are meant to incentivize the development of RL for efficient HVAC control do not generalize between buildings.",
"strength: The fidelity of the proposed (... | 2.5 |
McfYbKnpT8 | [
"strength: Novel Area of Research: The paper investigates a novel and promising application of deep learning to improve MIP solving, a problem that has not been significantly explored before.",
"strength: Framework Development: The authors have developed a new framework to systematically address the identified re... | 6.5 |
RlpJmARXqj | [
"strength: Studying on-device LLM personalization has significant practical application significance.",
"strength: The author claimed that the proposed method can achieve real-time model fine-tuning.",
"strength: The paper focus on decreasing computation as personalizing based on LLMs on-device.",
"strength: ... | 3.5 |
Hz4BYVY8YM | [
"strength: Novel Technical Contribution: The paper introduces a semi-automated pipeline that combines LLM-assisted generation with human verification to create temporal multi-turn QA chains, representing a methodologically sound approach to dataset creation.",
"strength: Comprehensive Empirical Validation: The ev... | 7.5 |
tdbK3TGFl1 | [
"strength: The paper proposes a new ranking problem with target \"documents\" linked within a directed acyclic graph.",
"strength: The proposed modification to a standard training of a dual encoder are simple.",
"strength: The paper is well structured and focuses on analyzing and understanding the limitations o... | 3.5 |
J9eKm7j6KD | [
"strength: The paper applies interpretability approaches to transformer-based models beyond the natural language domain and propose neural collapse as a metric of interpretability.",
"strength: The authors use sparse autoencoder-based steering for improving control vector linearity for motion control and zero-sho... | 4.8 |
4sJJixGIZX | [
"strength: One of the paper's main strengths is the formal introduction of the Online Continual Graph Learning (OCGL) framework.",
"strength: The authors develop a benchmarking environment specifically for OCGL, including multiple datasets and evaluations of various continual learning methods.",
"strength: The ... | 5 |
pQJi9EsmCc | [
"strength: This work proposes a direct treatment on foreground/background segmentation in Gaussian splatting framework, which is more integrated than the prior works depending on external cues such as SAM.",
"strength: The proposed method can start with a randomized initialization, which further get rid of depend... | 4.75 |
KjOAHlKMF5 | [
"strength: This paper studies a natural extension of a well-studied problem. Moreover, as summarized above, the contributions of this paper are clear.",
"strength: The main results of this paper, summarized in Section 4 (efficient oracle), Section 5 (regret minimization), and Section 6 (best policy identification... | 7 |
s9zoyICZ4k | [
"strength: The problem is well motivated. It is important to be solve different variants of VRP problems.",
"strength: From the numerical results, it appears that with the proposed approach, the accuracy of modeling is improved substantially.",
"strength: The work combines an algorithmic solution with queries t... | 6.5 |
ylgg2RE7ub | [
"strength: the paper is easy to understand.",
"strength: the results show the effectiveness of proposed pipeline.",
"strength: A novel approach to handling dynamic scenes without requiring traditional SfM-based initialization, and proposes to use the separation of static and dynamic components for independent o... | 4 |
RSGoXnS9GH | [
"strength: The paper contributes a novel fairness benchmark specifically for multi-turn dialogues, while current benchmarks primarily focus on single-turn dialogues.",
"strength: The paper extensively benchmarks on most popular LLMs, and provides detailed results and analysis across many dimensions, like tasks, d... | 7 |
VmGRoNDQgJ | [
"strength: Backdoor attacks for semantic segmentation models are an interesting threat model, apparently underexplored in prior works, and the paper fills this gap.",
"strength: The proposed methods are effective in the experimental evaluation on several architectures and datasets, and even in the real-world scen... | 7.5 |
kwqhn2VuG4 | [
"strength: The paper presents OmniCorpus, a large multimodal (text and vision) and multilingual (English and Chinese) dataset containing hundreds of millions of documents (bilions of images, and trilions of tokens). This is by far the largest publicly available dataset that I know of, which can increase the amount ... | 7.5 |
n7Sr8SW4bn | [
"strength: The authors' integration of Mayfly with machine learning is interesting.",
"strength: Well-written.",
"strength: The paper studies an interesting topic on using neural data structure for graph stream summarization, and designed the approach very neatly. In general, the presentation of the methodology... | 7.333333 |
o3pJU5QCtv | [
"strength: The proposed method leverages object-centric representations to generate action-state sequences for Behavior Cloning and shows that it generalizes to manipulating multiple objects. Experiments are conducted across three different tasks.",
"strength: This idea of predicting object manipulation actions t... | 6.25 |
3AAXabeZPG | [
"strength: This paper provides novel insights into visual and textual biases important for report generation and attempts to reduce them.",
"strength: It offers a thorough examination of these biases and their impact on model performance.",
"strength: The clear problem definition contributes to a more focused d... | 4.5 |
9ccZzuix2D | [
"strength: The experiments in this are comprehensive and sufficient.",
"strength: The theoretical motivation is well-written and reasonable.",
"strength: The paper is easy to read and follow.",
"strength: The paper is well-organized and easy to follow.",
"strength: Some empirical observations are intriguing... | 5.333333 |
LyJi5ugyJx | [
"strength: This is a very strong paper in analysis, practical techniques, writing, and experiment results.",
"strength: It studies an important but less studied problem: consistency models in continuous time, together with the training stability and discretization error of consistency models.",
"strength: The p... | 9.2 |
PabAln0jjB | [
"strength: Addressing temporal preference drift is an important and interesting problem.",
"strength: The paper is well written.",
"strength: The work provides both theoretical and experimental results.",
"strength: The paper is well-presented, highly readable, and organized effectively.",
"strength: The th... | 4.666667 |
L1p3uQ8pzl | [
"strength: The paper is well-written and well-organized.",
"strength: The proposed method improves the performance of LLMs in some pokemon tasks.",
"strength: It is interesting to discuss that the outworld context can let LLMs fail.",
"strength: The paper presents an interesting phenomenon that the knowledge ... | 6.2 |
vogtAV1GGL | [
"strength: The authors provide a novel formalism of \"concepts\", which facilitates further studies on \"concepts\".",
"strength: The authors provide proof for every statement they propose. They also provide examples to illustrate some abstract ideas.",
"strength: The mathematical treatment of the proposed fram... | 5.75 |
CscKx97jBi | [
"strength: There has been a lot of interest in LLM-based code generation in the research community recently. This work investigates an interesting and relevant problem.",
"strength: The idea of mimicking human debugging to improve code generation is interesting.",
"strength: In addition to regular experiments o... | 3 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.