paper_id stringlengths 10 10 | items listlengths 4 82 | gt float64 1 9.5 |
|---|---|---|
uHLgDEgiS5 | [
"strength: The method introduces a novel concept by capturing data influence in a trajectory-specific manner rather than assuming permutation invariance, which is a common limitation in conventional influence estimation methods.",
"strength: It outlines assumptions and derives an approximation error bound, lendin... | 8 |
MVltEnKJaO | [
"strength: Few-Step Generation: ASFM enables high-quality image generation in just a few steps, which is more efficient compared to traditional methods that require a large number of steps.",
"strength: Improved Efficiency: By leveraging straight flows, ASFM reduces the number of function evaluations (NFE) needed... | 4.75 |
4znwzG92CE | [
"strength: This work complements existing works (e.g. Habitat, VirtualHome, etc.) in a human-centric way. It allows more flexible human models, human-object interactions, human-robot interactions, and human-in-the-loop evaluation, which are often ignored in previous works.",
"strength: This work is in general wel... | 6.666667 |
KRqMfdwQaP | [
"strength: This paper addresses an interesting problem of improving the plausibility of output pose. The SEAL loss is new and effective.",
"strength: This paper is easy to understand.",
"strength: The problem of human pose lifting is of interest to the research community.",
"strength: Applying the SEAL framew... | 3.5 |
oZtt0pRnOl | [
"strength: The paper studies simple approach of achieving privacy in ICL (synthetic data generation and use it as demonstrations)",
"strength: Paper for most parts is clearly written and is easy to read.",
"strength: The empirical evaluation (including an ablation study) is very thorough.",
"strength: Compare... | 8 |
vAoyZWyDEc | [
"strength: The paper attempts to study lower bounds for nonconvex optimization. It claims that its contributions are to give negative answers to the following questions about any continuous nonconvex function $f$.",
"strength: The paper deals with computability issues in optimizing non-convex functions. This prob... | 2.5 |
dgmcE0RsTi | [
"strength: This paper shows that sparsification in cross-attention has a surprising potential for performance improvement, the proposed method improves about 20% over baselines.",
"strength: The authors propose a simple framework that can directly be used on existing PLMs for processing long sequences.",
"stren... | 5.5 |
fx8AJDQRVB | [
"strength: Appealing visual results are obtained with the proposed method.",
"strength: The paper is well-written and organized, making it easy to understand the proposed framework and its contributions.",
"strength: The author introduce sampling-space MOE to improve the image quality of diffusion-based SR, and... | 4.25 |
N4rYbQowE3 | [
"strength: The part on skip lists is a complete and meaningful contribution.",
"strength: The idea of promoting higher frequency elements to higher levels with more probability is both natural and easy to implement.",
"strength: Optimality as well as robustness of the proposed data structure is shown, and it is... | 7 |
qPwQj4Mf3u | [
"strength: Visual comparison of retrieved images and target images are provided via images, cosine similarity histograms, and loss.",
"strength: Cross-stimuli associations are considered.",
"strength: A wide range of temperature values was considered.",
"strength: The proposed idea in the paper is easy to fol... | 3 |
Mfnh1Sqdwf | [
"strength: When I compare this paper with other benchmarks (except EPInformer) in the field and for the same problem definition - I would rate this paper as high in terms of originality & significance. Yes, this work shows that gene expression prediction is not a siloed process, and that there are causal relationsh... | 7.5 |
7WAMJsDNDE | [
"strength: While a two-server model is not new, I like the idea that the proposed method introduces the dual server model to protect against MIA by preventing either server from accessing the final aggregation result.",
"strength: This paper introduces SHC, which allows users to verify aggregation correctness wit... | 4.5 |
STUGfUz8ob | [
"strength: This paper studies an important problem of generalization to unseen symbols in transformer architectures.",
"strength: Much of the contribution is formalizing a set of symbolic reasoning tasks which can be analyzed in different architectures.",
"strength: The authors give a theoretical analysis that ... | 7.6 |
auUngos7eR | [
"strength: The question that the paper attempts to address, namely online learning of state space models, is important in a variety of contexts and the authors offer a concrete pathway to do so, via their proposed optimization framework.",
"strength: The paper provides a potential novel direction that can be foll... | 6.666667 |
iVxxgZlXh6 | [
"strength: The pursuit of robotics-oriented pretext tasks for training large-capacity open source models remains quite compelling.",
"strength: The paper reformulates BC data into an image-text format compatible with VLMs. This is not a new idea (RT-2), but they do introduce a new output space and auxiliary objec... | 5.25 |
5IkDAfabuo | [
"strength: This work proposes a scalable method for training model-free or model-based agents in a variety of domains. I believe the formulation is simple enough to be integrated into and improve other approaches.",
"strength: I also found the presentation clear and easy to read.",
"strength: I found the scalin... | 7.5 |
neDGc4slhd | [
"strength: The proposed research direction of comparing functional circuits using topological tools is interesting",
"strength: Novel Approach: Using TDA tools to analyze neural networks across datasets and epochs is an intriguing approach, especially in trying to capture global behavior via Betti curves.",
"st... | 2.857143 |
1S8ndwxMts | [
"strength: Unifying benchmarking attempts for protein generative models is an extremely important open challenge.",
"strength: Studying various common evaluation metrics systematically and in a controlled setup is impactful because it can inform future developments of new methods and allow researchers to benchmar... | 3 |
MSe8YFbhUE | [
"strength: The insight is derived from seemingly unrelated but recent work on Neuron Dormancy, which explains how the phenomenon can impact learning in neural networks through a potential plasticity mechanism. This paper seems to provides a novel lens for the phenomenon, showing that neural activity can be correlat... | 6.5 |
gaa7gWPZBz | [
"strength: The paper tackles an important topic in the field of machine learning, specifically the tradeoff between privacy and performance in the context of adversarial examples.",
"strength: The authors propose an approach to generating counterfactual explanations that are aligned with the original data distrib... | 3 |
dWsdJAXjQD | [
"strength: This paper first discovers a new field of automated proof optimization, and conducts a pioneering attempt in this field, marking the first significant effort to use LLMs for enhancing formal proofs with respect to certain metrics in proof assistants like Lean.",
"strength: A notable strength of this wo... | 6.75 |
oOa3ZCtMjJ | [
"strength: This article is well-written and easy to understand.",
"strength: It leverages the semantic alignment strengths of CLIP to enhance the quality of images generated by GANs.",
"strength: By analyzing text to identify entities, attributes, and relationships, the model can generate images that are not on... | 3 |
UYXq4q1GpW | [
"strength: It is nice to consider the characteristics of food items, i.e., the food nutrition, when recommending the food items to users.",
"strength: The paper tackles an interesting aspect of food recommendation, namely, the introduction of healthy suggestions alongside the usual accurate ones",
"strength: Th... | 2 |
iGHPVbttMs | [
"strength: The authors provide insights into the behavior of a large class of empirical game theoretic algorithms, and use the insights to provide improvement on the state of the art.",
"strength: The strength of the result in the manuscript is indeed noteworthy as it provides a contemporary interpretation and fo... | 3.4 |
dWReNWEj5b | [
"strength: The use of adapters allows for parameter-efficient training, and easy accommodates various types of conditioning information, such as enzyme function and taxonomy for different applications.",
"strength: ProCALM shows capabilities to generalize to rare and unseen enzyme families and taxonomies, which i... | 5.75 |
f6GMwpxXHG | [
"strength: Originality and significance: To the best of my knowledge this is the first work to propose this type of GAN training objective. Improving stability of GAN training by increasing robustness properties via losses more robust to outliers and with bounded gradients is an interesting idea worth investigating... | 2.2 |
gc8QAQfXv6 | [
"strength: The forgetting issue during fine-tuning foundation models is an important problem.",
"strength: The observations and analysis through the function vector in this paper is very interesting, which I believe would be interesting to many.",
"strength: The finding that CF in LLMs primarily stems from bias... | 9 |
1iuaxjssVp | [
"strength: An interesting idea to approach the inverse folding problem (i.e. the problem of generating sequences that fold into a given structure).",
"strength: Proposes a low-rank approximation of the couplings and fields of the lightweight sequence model.",
"strength: Fast generation of sequences that fit a w... | 7.25 |
51WraMid8K | [
"strength: Authors provide 4 carefully defined metrics, along with necessary guarantee proofs. Overall, the paper is very well composed.",
"strength: Those 4 proposed metrics allow to comprehensively evaluate model unlearning using entire output distribution (potentially, right now it is limited to MC sampling on... | 8 |
hPk92D2GJV | [
"strength: This paper presents FaithfulBench, encompassing the tasks of Openbook QA and Multi-choice QA based on the NaturalQuestions and TriviaQA datasets. Experimental results show that all models exhibit a degree of stubbornness to their internal knowledge, and there exists a significant negative correlation bet... | 5.25 |
qpXctF2aLZ | [
"strength: The motivation is good, aiming to improve the interpretability of RL, which is a key factor for deployment due to its adoption of DNNs.",
"strength: The authors try to improve the stability of DTs as a policy via more implemented skills.",
"strength: The proposed method can be applied into both discr... | 7.2 |
0jmFRA64Vw | [
"strength: The paper reports and discusses the empirical performances of FedComLoc in different configurations and hyperparameter settings, including heterogeneity, quantization bits, and TopK sparsities.",
"strength: The paper presents some interesting observations based on the numerical results.",
"strength: ... | 3 |
MUL7tKvNei | [
"strength: The motivation of this paper is clear.",
"strength: The three modules make sense to me.",
"strength: The authors have created numerous figures and visualizations to clearly illustrate the algorithm framework and experimental results.",
"strength: The proposed approach appears fresh compared to exis... | 4 |
BhECSDSkAE | [
"strength: The paper tackles an important problem of generalizing dense segmentation task over videos using sparsely annotated frame (in their case the first annotated frame).",
"strength: The experiments on several medical videos show that their method outperform different one-shot, cross-domain one-shot and tes... | 3.666667 |
RDLvnUJ5JZ | [
"strength: The paper effectively adapts diffusion models for time-series forecasting, a relatively novel domain for such models, building on their successes in generative tasks.",
"strength: The authors integrate existing diffusion models for time-series forecasting into a continuous score-based framework, enhanc... | 3 |
ztzZDzgfrh | [
"strength: Authors provide a straightforward method to detect hallucinations in RAGs that does not require model fine-tuning.",
"strength: Empirical results provided by the authors look good.",
"strength: Each step is thoughtfully motivated, with both conceptual reasoning and empirical validations in §3. The de... | 7.333333 |
6cMmSnOpCs | [
"strength: The paper is clearly written, and is generally easy to follow.",
"strength: The method is simple, well-motivated, and works well.",
"strength: In my opinion, the method proposed in this paper should have been a baseline in the original [Pfeiffer et al](https://arxiv.org/abs/2005.00247) paper that int... | 5.75 |
SUc1UOWndp | [
"strength: The rLLC provides an innovative, quantitative tool for tracking specialization in Transformer models, which could have applications beyond this study.",
"strength: The paper includes a comprehensive analysis of attention head specialization, backed by clustering and ablation studies that demonstrate th... | 7 |
5BjQOUXq7i | [
"strength: The paper introduces a novel, regression-based approach for selecting data mixtures that reduces computational costs in language model training. The approach offers an efficient alternative to traditional dynamic or heuristic data allocation methods, making a valuable contribution to the field.",
"stre... | 7.2 |
cINwAhrgLf | [
"strength: The paper is well-written with clearly motivated arguments and insights. The auxiliary learning task is also meaningful when we only seek to boost one task with another and aim at quick inference.",
"strength: Table 1 provides a comprehensive understanding and meticulous survey of the field. The author... | 7.2 |
tccML2tDd4 | [
"strength: Good idea to apply human visual cues to object detection in foggy scenes.",
"strength: Clear figures to show the proposed overall architecture of Perceptual Piercing and the AOD-NetX modules, making it easier for readers to understand, which improves the quality of the presentation.",
"strength: In g... | 2.333333 |
n4SLaq5GhM | [
"strength: The paper nicely overviews overarching weaknesses in the current generation of AI + NLP technologies and electronic health record (including bias, data fragmentation, missingness, automation bias).",
"strength: The authors outline in detail possible limitations of their proposal.",
"strength: The nov... | 3.25 |
bJx4iOIOxn | [
"strength: The paper discusses an interesting and important subject, and gives sensible and valuable insights on the topic.",
"strength: The whole analysis is thorough enough and the conclusions seem credible.",
"strength: The paper is overall well written and easy to follow.",
"strength: The paper is well-wr... | 7.5 |
Kn7tWhuetn | [
"strength: The paper is well motivated and clear. The proposed changes are supported by the empirical evaluation on the CLRS-30, outperforming the baseline in most algorithms.",
"strength: for better generalization, the authors proposed to restrict the function class by excluding non-Markov solutions, leading to ... | 6.8 |
Ad81awoBVS | [
"strength: The method introduced detects rotations that remain unchanged and are semantically similar to original images, ensuring their exclusion as outliers. This method demonstrates sufficient generalizability for other transformation (ex., gaussian noise), as supported by section 4.3. I feel that this contribut... | 5.4 |
Zz61cEY84L | [
"strength: The authors present a learning effort framework over a number of problem settings to analyze optimal strategies for learning. Having an understanding and intuition about this can help our current deep learning design problems learn better.",
"strength: The document motivates the problem well, emphasizi... | 6.25 |
zp88xOXAfS | [
"strength: LICEM offers a transparent, linearly interpretable structure, allowing users to understand how different concepts contribute to predictions.",
"strength: The model maintains accuracy comparable to black-box models while enhancing interpretability.",
"strength: A Self supervised LICEM variant reduces ... | 4.8 |
fw1oizreEF | [
"strength: The paper is well-written and easy to read.",
"strength: Theoretical derivations are easy to follow.",
"strength: The proposed convex alternative to self-attention presents a natural way to interpret the mechanism. It is interesting to see that an implicit regularization on token sparsity shows up af... | 5 |
ynguffsGfa | [
"strength: The paper proposes an intuitive and well‑motivated approach to leverage background knowledge of GPT‑4 for informed tabular data generation",
"strength: The experimental results are promising and cover a representative set of baseline tabular data generation methods",
"strength: The idea of generated ... | 6.333333 |
jsQPjIaNNh | [
"strength: The framework is equipped with mathematical foundation and its effectiveness can be theoretically understood.",
"strength: Understanding the function of unseen proteins is a really important task in protein engineering and other biological applications.",
"strength: The paper is well-written and easy... | 5.25 |
IEKQlWIN4w | [
"strength: The writing is easy to follow.",
"strength: The motivation and proposed method are clearly stated.",
"strength: This work represents a novel effort to integrate multi-agent systems into Auxiliary Modality Learning.",
"strength: The experiments demonstrate the effectiveness of CAML compared to previ... | 5.25 |
CZiY6OLktd | [
"strength: The research problem addressed in this study is of paramount significance and holds great interest. Accurate time prediction has broad applications, including tasks like anomaly detection and energy consumption control.",
"strength: The paper introduces a novel and intriguing approach by linking variou... | 6 |
cSd8Eom8Zt | [
"strength: The method shows some promise in improving the accuracy of the classification models and produces good clustering performance.",
"strength: The only strength of the submission is in the core motivation of finding better representation of Deep Networks that would match strong prior practitioners would h... | 2.333333 |
xgtXkyqw1f | [
"strength: The paper demonstrates considerably better output responses for MindSearch, compared to proprietary AI-Search engines like Perplexity Pro and ChatGPT-Web.",
"strength: MindSearch also works considerably better than the closed-book and ReACT baselines on a variety of multi-hop question-answering dataset... | 5.75 |
YrxhSkfHh0 | [
"strength: The paper's readability is generally good.",
"strength: It provides a nice background behind the development of HGR and clearly states that its goal is to make Soft-HGR more computationally feasible.",
"strength: Experimental results on MNIST, Houston2018, IEMOCAP compared to CCA, Deep CCA, Soft CCA,... | 3.333333 |
oEzY6fRUMH | [
"strength: This paper studies an interesting problem of learning effective representations for reinforcement learning",
"strength: This paper proposes a variety of different losses for temporal representation learning",
"strength: As the authors point out, capturing long-term behaviors is a key issue in RL.",
... | 4.75 |
hfAEEsIQ6D | [
"strength: The proposed variations over existing approaches sound interesting and grounded on human cognition.",
"strength: By building upon the foundations laid out by prior research, this paper takes strides in refining and enhancing what has been previously suggested in the domain.",
"strength: It is a simpl... | 4.25 |
dsHpulHpOK | [
"strength: The paper is well written, and for me, it was easy to understand, in addition, it cites many relevant papers across all fields (RL, control theory, medicine).",
"strength: I think this paper bridges some of the gap between real-life problems to RL; while it is more applicative in terms of RL, it sugges... | 7.5 |
hbon6Jbp9Q | [
"strength: The paper is certainly well motivated - understanding the feature dimensions that underlie brain predictivity is a core open challenge in cognitive neuroscience.",
"strength: I commend the authors for pursuing an approach that is fundamentally data-driven, avoiding prespecified hypotheses about the inv... | 2.333333 |
sPuLtU32av | [
"strength: Theoretical contribution: The framework introduces a structured approach to sparsification in training, with rigorous theoretical analysis of convergence for both convex and non-convex cases.",
"strength: Relevance: This paper addresses a very relevant topic in the field.",
"strength: Scalability: Ex... | 7 |
M7KyLjuN0A | [
"strength: Compared to existing methods that lack the ability of long-term dynamic generation, this paper utilizes Hexplane as the compact 4D representation and reorganizes into one feature map to achieve efficient reconstruction and generation.",
"strength: The decoding manner in parallel proposed in Expansion &... | 7.5 |
Ry1SZkcYbX | [
"strength: The paper is well-motivated, which indicates the time-consuming issue caused by the large sample counts on sign distance function (SDF) thin edges while employing error-bounded sampling.",
"strength: The manuscript is easy to follow, with all concepts clearly explained.",
"strength: A fast, accurate,... | 3.5 |
pRCTRC6icM | [
"strength: The target problem of data efficiency is important and has the potential for broad applicability.",
"strength: The proposed method is simple yet highly effective. Its effectiveness is thoroughly validated through extensive experiments across various components, hyperparameters, datasets, and architectu... | 5 |
SO0manOwUF | [
"strength: The concept of using uncertainty quantification (UQ) to guide the merging of multimodal large language models (MLLMs) is relatively novel.",
"strength: The paper is generally well-organized, with clear figures illustrating the architecture and merging process, as well as tables comparing the performanc... | 5.5 |
aIAFDFpNXz | [
"strength: The system architecture is well thought of. It uses LMMs to generate code functions as semantic-level skills, which encapsulate lower-level keyboard and mouse control.",
"strength: Codebase has been shared and the appendix is extensive with figures and supporting material, case studies. Hard work has b... | 6.5 |
sahQq2sH5x | [
"strength: Addresses Core Issues in PC Research: The paper tackles fundamental shortcomings in the subfield, such as inconsistent comparisons, overly simple benchmarks, and a lack of rigorous, standardized practices.",
"strength: Unified Benchmarking Framework: Introduces a comprehensive and standardized benchmar... | 7.333333 |
NgtEafc8NZ | [
"strength: The motivation for this work is easy to understand. The authors straightforwardly introduce the problem they are trying to solve and explain why it is important.",
"strength: The background material is abundant and comprehensive, crediting the proper related works.",
"strength: The experiments demons... | 5 |
THUBTfSAS2 | [
"strength: At least in some parts of the paper, the flow is very good, and theorems and conclusions naturally lead to the next part.",
"strength: The evidence for diversity in LDM-based active learning, as discussed in Appendix E.3, is extraordinary. Through an intuitive first example and then a real-world exampl... | 5.25 |
7JigPd5Pm5 | [
"strength: The paper introduces a theoretically informed weight initialization method to mitigate oversmoothness, which is a novel departure from existing approaches.",
"strength: The idea of informed initialization is important for any neural network, and also for graph neural networks, so the paper considers an... | 2.5 |
Ts95eXsPBc | [
"strength: Great Motivation: The paper addresses a critical gap in existing models’ inability to effectively integrate spatial information, which is very crucial for tasks in various domains. The introduction and literature review provides a compelling argument for why this integration is necessary, setting a solid... | 7 |
BxLK1M1f8T | [
"strength: The core idea of the paper is interesting and well-motivated, addressing the challenge of aligning predicted actions with desired returns in offline RL.",
"strength: The paper is well-written and easy to follow, making it accessible to readers with a background in RL and transformers.",
"strength: Th... | 5 |
GM7cmQfk2F | [
"strength: The method of deploying \"conditional attention\" as proposed in the paper is simple and novel.",
"strength: With the exception of the points discussed in the Weaknesses Section, the paper is of good quality.",
"strength: The paper features a comprehensive list of experiments. It discusses variations... | 7 |
fMTPkDEhLQ | [
"strength: The problem studied in the paper is both important and well-motivated.",
"strength: The paper provides a rigorous analysis with novel contributions to the study of oracle complexity for asymmetric high-order convex functions, extending previous research by offering a general framework.",
"strength: T... | 8 |
iTFdNLHE7k | [
"strength: I commend the authors for their dedicated focus on tackling the challenge of low-data scenarios in generative modeling.",
"strength: I value their efforts in quantifying the improvements brought about by their proposed architectures, particularly in terms of reducing computational demands for hyperpara... | 6.75 |
b2UlHeyyC0 | [
"strength: The authors have thoroughly investigated various designs for retrieval enhancement, emphasizing the importance of combining uni-modal search and cross-modal fusion.",
"strength: The proposed RECO employs a light-weight, single-layer transformer encoder for fusion, without significantly increasing the n... | 5.666667 |
c2C2NQKjZw | [
"strength: Nice to see a benchmark inspired by actual use",
"strength: Neat trick to generate test cases to assess code completion quality",
"strength: *Data-driven design*: The benchmark is grounded in insights from an industrial code completion tool, allowing Codev-Bench to closely align with real-world devel... | 4.25 |
Ur4LqAOXIF | [
"strength: The paper is simple and easy to follow and understand. It gently introduces the problem and smoothly translates to the proposed method.",
"strength: The proposed algorithm (SODA) is simple and neat.",
"strength: The provided experimental results shows the effectiveness of the proposed SODA compared t... | 3.5 |
H2Gxil855b | [
"strength: The paper is well-written and organized; hence it is easy for readers to follow.",
"strength: Addressing the current challenges of VAEs in extracting efficient latent representations for generative diffusion models is important.",
"strength: The introduction of the Atlas Gaussians representation with... | 7 |
MrR3rMxqqv | [
"strength: The paper makes theoretical contributions by exploring the memorization capacity of transformers, an area that is not yet fully understood. This contributes to a deeper understanding of transformer architectures.",
"strength: The paper introduces new assumptions about the linear independence of input d... | 7.5 |
T7ZVzuObcj | [
"strength: The paper is well-written and easy to follow.",
"strength: It claims to be the first work to achieve locally interpretable point cloud classification on a per-point basis.",
"strength: The visual results are compelling, highlighting important regions for classification. Quantitative results on multip... | 5.5 |
ajSmXqgS24 | [
"strength: Meticulous and concise description of the details of the proposed algorithm.",
"strength: Tackles an important problem of learning a tracking controller for dexterous manipulation that is often overlooked, but could have significant impact as our data collection devices continue to improve with the rap... | 6.25 |
KIgaAqEFHW | [
"strength: The \\texttt{miniCTX} benchmark fills the gap in the current community: it expands current benchmarks by addressing the limitations of standalone theorem proving and enabling evaluation in real-world scenarios where context is critical.",
"strength: I appreciate the authors' commitment to automatically... | 8 |
lvjz7Bm3Ea | [
"strength: This is an applied paper and has little technical contribution.",
"strength: The experiment is relatively comprehensive on 85 benchmark datasets and derived 652 datasets.",
"strength: Simple and efficient reminiscences of numerical challenges in One-Class-Learning.",
"strength: Extensive experiment... | 3.75 |
ExiBN1ZWJn | [
"strength: This is an interesting topic to apply DDPM on graph representation learning.",
"strength: The authors had some good ideas on using Laplacian smoother and a coupled node feature similarity based edge removal schedule to add noises.",
"strength: They claimed this helps learn a more general representati... | 4.75 |
xjornbs7aT | [
"strength: The motivation for this work is clear: it addresses the SCMDP problem by utilizing a model to learn a feasible action space, effectively converting SCMDP into an MDP and improving the algorithm’s sample efficiency.",
"strength: The paper provides a step-by-step explanation of related concepts, making i... | 4.25 |
IcNzKiB8CP | [
"strength: The idea to utlize (layer) input-space null projection is built atop the work of GPM (Saha et al'21) in continual learning and its use here for the purposes of teleportation is neat.",
"strength: This extends the applicability of the approach and might draw further interest in these methods. The result... | 3.75 |
Gf1uBeuUJW | [
"strength: The paper introduces temporal hacking as a new perspective on reward hacking in RL and effectively adapts this concept to video MLLMs.",
"strength: TPL score is a notable contribution, providing a quantitative way to assess temporal misalignment, a challenge often overlooked in video modeling.",
"str... | 6.5 |
jbfDg4DgAk | [
"strength: By using POS tags to select tokens for watermarking, SpARK ensures that watermarking is applied selectively and minimally, reducing the model's interference with the natural language structure.",
"strength: The methodology section is well-structured, with algorithms (e.g., POSWatermark, DetectWatermark... | 3 |
3i13Gev2hV | [
"strength: The proposed method is simple and elegant and can be easily applied to large scale pretraining of vision-language models. The procedure to automatically generate paired image and text boxes is also relatively straightforward.",
"strength: The empirical results show improvement across several tasks whic... | 8 |
ljwoQ3cvQh | [
"strength: Finding of the paper is interesting.",
"strength: Paper writing is careful and clear.",
"strength: The paper includes detailed evidence, both empirically and theoretically, for their claims.",
"strength: [Originality] The paper novelly reassesses DNN’s OOD behavior both empirically and theoreticall... | 7 |
dLrhRIMVmB | [
"strength: Understanding potential applications of NISQ is a very important question in the entire field of quantum computing. The work provides a significant step toward obtaining an end-to-end application of NISQ by proposing a NISQ algorithm for topological data analysis.",
"strength: Topological data analysis... | 8 |
ImwrWH6U0Y | [
"strength: The paper are well written and the problem is well-motivated.",
"strength: The experiment are comprehensive and some empirical findings are insightful (e.g., CL cause disparate impact to members and non-members is useful)",
"strength: This paper is the first work to study the privacy risks introduced... | 4.75 |
N5fVv6PZGz | [
"strength: The paper systematically approached scenarios where mixed use of CPUs and GPUs can outperform performance of purely relying on GPUs for MoE models.",
"strength: The algorithm in the experiment achieved good results.Twiddler achieves 1.26 times speed up in single batch inference, 1.30 times in long pref... | 6.333333 |
GbEmJmnQCz | [
"strength: Due to the interest in F&Z---the article currently has 426 citations---undertaking this sort of investigation seems quite worthwhile, and such careful analyses of others' work is sadly uncommon in our field.",
"strength: The paper is generally well written and clear.",
"strength: The experiments are ... | 4.4 |
0zGvf2yRMQ | [
"strength: The final images do indeed look better than the rendered comparison images, for the images shown.",
"strength: The paper is easy to read.",
"strength: The results in all figs look good, when comparing with existing methods, especially for the geometric details.",
"strength: Authors provide a compre... | 5 |
o1efpbvR6v | [
"strength: The use of Attention mechanisms and Proxy Anchor Loss enhances the performance of classification tasks by capturing both local and global contexts.",
"strength: The research effectively addresses a practical challenge in the field - the imbalance of template distributions in datasets.",
"strength: Th... | 2.333333 |
pCx6DYN43D | [
"strength: It introduces a unique tokenization method of using visual tokens to represent textual content, effectively compacting long-context data and allowing for efficient processing within a limited token length, which leads to faster inference times.",
"strength: SEEKER outperforms existing multimodal models... | 4.333333 |
Giwj9cgAIl | [
"strength: The work provides a very interesting methodology a single governing equation of a linear/nonlinear system in an interpretable manner.",
"strength: The authors show through their experiments that they are able to learn nonlinear ODEs (unlike previous work like SINDy).",
"strength: The methodology also... | 4.666667 |
owziuM1nsR | [
"strength: The idea of using global attention in the Transformer is widespread. But, the authors effectively maintain low computational complexity, which is meaningful in image SR.",
"strength: Additionally, the proposed HAI is simple yet effective. Both the ablation experiments (Table 1 (c), (d)) and the visual ... | 7.5 |
C7XoUdJ5ZC | [
"strength: The experiment section of the work shows considerable improvements that FLAIR can achieve over prior work, especially in more extreme non-IID scenarios.",
"strength: The authors mix two key methods in machine learning to develop their solution (i) Reptile meta-learning, and (ii) CVAE.",
"strength: Ad... | 3 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.