paper_id stringlengths 10 10 | items listlengths 4 82 | gt float64 1 9.5 |
|---|---|---|
m2nmp8P5in | [
"strength: Innovative approach: It is innovative to use LLMs for scientific equation discovery and symbolic regression. LLM-SR integrates LLMs' prior scientific knowledge and and code generation ability to achieve good performance.",
"strength: Robust evaluation: The authors have carefully designed benchmark prob... | 8 |
4Hf5pbk74h | [
"strength: The proposal of analyzing the latent space through the lens of nearest neighbors makes sense and can provide some insight on model behavior.",
"strength: Results reported show small, although noticeable performance improvement over the baseline models.",
"strength: The proposed method is simple and c... | 2.333333 |
0dELcFHig2 | [
"strength: This work is novel in that it is the first to use fMRI. But other works also take similar approaches to identifying multimodal processing (see weaknesses).",
"strength: Findings of the difference between cross-modal and jointly-trained models with respect to regions is novel",
"strength: These findin... | 6.666667 |
iuxaCU3DI7 | [
"strength: The paper is well presented and easy to follow.",
"strength: Although the technical contribution is limited, the efforts and contributions in the paper are great and are likely to help the research community in the field of surgical tool detection, segmentation and analysis.",
"strength: The evaluati... | 7.5 |
hBGavkf61a | [
"strength: The paper addresses an important issue in representation learning with diffusion models.",
"strength: The proposed solution appears theoretically sound and empirically convincing to me, but I have to admit that I am not an expert on the relevant literature regarding other solutions to the discussed iss... | 7.25 |
e9yfCY7Q3U | [
"strength: The field of LLM jailbreaking is currently the focus of the LLM security community and is central to developing secure LLM-based products. As such, the research is timely and interesting.",
"strength: Authors perform an extensive comparison of their method not only to the original method but also to ot... | 6.25 |
Wr3UuEx72f | [
"strength: This paper introduces a novel method for video tokenization that moves beyond traditional patchwise encoding, making it both interesting and innovative.",
"strength: The paper provides a comprehensive set of experiments, including video reconstruction, class-conditional video generation, and video fram... | 7.5 |
tIsYpRxMvr | [
"strength: This paper employs a diffusion model, which has shown promising performance in generative model training. The idea of this paper seems to be novel.",
"strength: The authors provide theoretical derivations and empirical results demonstrate good results.",
"strength: provide the proof of predicting how... | 4.25 |
04RGjODVj3 | [
"strength: This study try to address the important issue of cross-user variability and BCI illiteracy issues in the MI-EEG analysis by adopting the ability of HyperNetwork to adaptive weight generation to learn user-specific representations.",
"strength: The paper is original in that they apply a HyperNet archite... | 3 |
5LhYYajlqV | [
"strength: The paper poses a very interesting problem and it can motivate future works.",
"strength: The paper is written nicely and it is easy to follow.",
"strength: The paper brings a novel approach to the domain of machine unlearning, effectively leveraging the In-Context Learning (ICL) paradigm to achieve ... | 5.333333 |
FwkYeLovHk | [
"strength: The idea of weak-to-strong generalization is somewhat meaningful.",
"strength: The efficiency is improved by removing the text encoder in the inference stage.",
"strength: This paper is the first to explore weak-to-strong generalization in CLIP-based classification.",
"strength: This paper proposes... | 3.333333 |
ZJHdiYDD5k | [
"strength: The propose method is well-motivated.",
"strength: The paper is well-structured, capable of clearly elucidating its core ideas.",
"strength: The conducted experiments adequately showcase the efficacy of the method being proposed.",
"strength: The motivation is well presented, the analysis of const... | 5.5 |
PqeMnyGU1B | [
"strength: The paper offers a simpler and more computationally efficient approach to improving SWD by introducing a single global rescaling factor based on slice informativeness.",
"strength: The rescaling approach can be absorbed into the learning rate tuning process, making it easy to implement without extensiv... | 6.333333 |
9uELGn17Db | [
"strength: Sampling with a slightly higher temperature energy surface and reweighting during learning to reduce the instability of negative samples is an interesting idea which seems to provide some stability benefits. The math behind the reweighting method is sound.",
"strength: The paper is presented clearly, w... | 3.5 |
zgHamUBuuO | [
"strength: Sparling achieves extreme activation sparsity and promotes learning interpretable representations.",
"strength: The algorithm shows good localization performance on the provided evaluation datasets.",
"strength: The idea of using activation sparsity to maintain representation capacity while allowing ... | 3 |
FLOaCQfZe9 | [
"strength: The idea of augmenting both tasks and trajectory data has not been explored in previous works and is technically challenging.",
"strength: Experimental results demonstrate promising sample efficiency and generalization capabilities compared to existing meta-RL methods.",
"strength: This work attempts... | 2.5 |
VAqRZIuW8m | [
"strength: Using mixture-of-experts for multi-task learning is novel and interesting.",
"strength: Experiemtal results on three datasets as well as the ablation studies show the effectiveness of the proposed approach.",
"strength: The motivation is clear and Figure 1 is helpful to understand the proposed method... | 3.5 |
EHmjRIA4l2 | [
"strength: This paper is written clearly and motivates the need for abstracted world models.",
"strength: The usage of hypernetworks for this task is novel and an interesting approach to solving the hierarchical problem.",
"strength: The authors also provide interesting ablative experiments of their method in g... | 3 |
agocj3HTTd | [
"strength: Interpretability is an important topic.",
"strength: The motivation is clearly demonstrated.",
"strength: The method is easy to follow.",
"strength: The target problem addressed by this paper is both interesting and crucial for advancing interpretability in federated learning.",
"strength: The pr... | 2.333333 |
swqZKDoMJA | [
"strength: Overall, the paper makes solid contribution to minimax optimization. The idea of Decoupled SGDA seems to be fairly natural, but at the same time, it means the simplicity of the algorithm could make it practical.",
"strength: Reducing the communication complexity of decentralized learning dynamics in ga... | 6.5 |
ewIfVxCzbo | [
"strength: The paper identifies an important problem that is not well-addressed",
"strength: The paper clearly shows two challenges and proposes effective solutions and the results look promising",
"strength: The experiments seem solid to support the claims on multiple sources and tasks.",
"strength: The pape... | 5 |
PkpNRmBZ32 | [
"strength: The concept of generalized SSM blocks, configurable with flexible connectivity structures, demonstrates good soundness and could complement existing SSM designs.",
"strength: The proposed method is supported by both theoretical analysis and empirical experiments.",
"strength: The presentation of meth... | 7.5 |
yV6wwEbtkR | [
"strength: In implementation sense, the idea looks simple and easy to implement this idea; introducing a estimate of the MCMI in Eq (2) is additionally necessary.",
"strength: It seems that the proposed objective for the teacher model can be integrated with existing knowledge distillation algorithms which mainly ... | 6.666667 |
1YSJW69CFQ | [
"strength: Addressing the reliability of segmentation systems in practice is a timely topic.",
"strength: To name one, the broad topic that this paper covers, i.e. uncertainty estimation in medical imaging, is an important one.",
"strength: This paper investigates the critical challenge of quantifying uncertain... | 1.666667 |
urf8a5G59f | [
"strength: The authors demonstrate superior performance in several benchmarking tasks, including brain tumor and full body MRIs.",
"strength: The authors claim that the proposed model is able to generalize to novel domains, not seen during the training.",
"strength: The problem, motivation and contributions are... | 5.5 |
Nz2UApmv2e | [
"strength: This paper showed novelty as a first implementation in multi-task learning with full SNNs in audio domain, while previous papers on SNNs have focused more on single classification task.",
"strength: Strong benchmarking performance in showing superior accuracy in both SEC and localization tasks, lower M... | 5 |
f4gF6AIHRy | [
"strength: Well-Motivated Research: The paper addresses a significant issue in data selection for LLMs—dimensional collapse due to lack of diversity—which has substantial implications for model performance.",
"strength: Clear Presentation: The methodology is clearly explained, making the paper easy to read and un... | 8 |
xlbXRJ2XCP | [
"strength: Introducing the concept of maxcut into attributed graph pooling is novel and meaningful, and it is worth noting the proposed method is feature-aware and differentiable.",
"strength: The background and related work are introduced in detail and accompanied by illustrations, which greatly aid in understan... | 5.25 |
J1SGf2lyr6 | [
"strength: This work offers a Feature-Aware Federated Learning framework for anomaly detection in 5G networks, addressing challenges of data privacy and scalability.",
"strength: The proposed framework effectively addresses conditional joint learning methods in the context of data heterogeneity, lack of labelling... | 2.5 |
rzF0R6GOd4 | [
"strength: This paper brilliantly bridge the concept of sceneflow and the proposed SDF flow.",
"strength: Based on the two assumptions, one for rigidity and the other for the linearlization (?), the proposed SDF flow is differentiably convertible into the sceneflow, which enables the network to encode sceneflow f... | 8 |
j0sq9r3HFv | [
"strength: Using LLMs to extract parameters for building SNNs is an interesting topic.",
"strength: The intuition is good.",
"strength: The project aims to tackle relevant research questions.",
"strength: The use of open source LLMs ensures wider applicability of the method.",
"strength: The conclusions doe... | 2.5 |
R1hIXdST22 | [
"strength: MR.Q combines various state-of-the-art techniques and demonstrates promising performance gains with a single set of hyperparameters across multiple benchmarks, showing both efficacy and resource efficiency.",
"strength: The paper provides detailed ablation studies, offering insights that may benefit fu... | 7.5 |
SL7djdVpde | [
"strength: A novel idea to use Dynamical Lie Algebra and over-parameterization theory for Parametrized Quantum Circuit (PQC) to design symmetry-preserving VQA ansatz.",
"strength: Very clear explanation of the theoretical tools used.",
"strength: The authors conduct extensive numerical studies showing positive ... | 6.75 |
KLUDshUx2V | [
"strength: Innovative Methodology: The study introduces an unsupervised concept generation method and a CLIP-based multimodal evaluation scheme, providing a more automated and adaptable solution for concept bottleneck models.",
"strength: Comprehensiveness: The experiments cover three different datasets and compa... | 3.4 |
CEkIyshNbC | [
"strength: The double descent is, currently, highly counter-intuitive to many, so any insights into the phenomena are valuable.",
"strength: The stated hypothesis regarding label noise is clear and seems intuitive.",
"strength: This paper is feasible to understand the generalization ability and double descent p... | 4.5 |
I9aemDuy5b | [
"strength: The $\\ell_p$ geometry has attracted a lot of attention in recent years due to the similarity of Adam to signSGD. The theory, however, is still quite limited and is usually restricted to the case $p=\\infty$, while the authors make an interesting observation that a different value of $p$ might turn out t... | 3.5 |
pdzHpQbGrn | [
"strength: Proposes a framework for prompt learning in vision-language models.",
"strength: Shows reasonable performance gains.",
"strength: The intersection of active learning and test-time adaptation is increasingly crucial, and the authors are timely exploring that field.",
"strength: This paper first appl... | 2.5 |
J7hbPeOZ39 | [
"strength: The paper is written clearly.",
"strength: The problem looks new.",
"strength: The algorithm performs good theoretically and numerically.",
"strength: The paper is well-written, addresses a practical problem, and offers a UCB-based algorithm with theoretical guarantees that matches the lower bound.... | 3 |
BkRD6GsswM | [
"strength: The approach taken by the authors is sensible, splitting the components of selecting an instance to be labeled and then selecting whom should do the labeling.",
"strength: The paper provides thorough citations.",
"strength: To address the issue of active learning with multiple annotators, the paper p... | 3.5 |
CbpWPbYHuv | [
"strength: (clarity) Overall, the paper is clearly written and easy to follow.",
"strength: (originality) The proposed activation functions are novel (to the best of my knowledge).",
"strength: PolyReLU and PolyNorm show nontrivial performance gains for training language models with >1B parameters, even compare... | 7 |
1Akd36hG9z | [
"strength: The proposed method is well-motivated, and using DiCE for weighting data is new.",
"strength: The theoretical results are encouraged.",
"strength: Theoretical analysis: The OSD-DICE framework presents a theoretically sound approach, advancing the field with its single minimization objective that reso... | 5.333333 |
1k4yZbbDqX | [
"strength: The results of this paper are truly captivating.",
"strength: It manages to generate images of impressive quality with just one or two steps.",
"strength: The quality showcased in the figures is highly satisfying.",
"strength: In addition, the paper provides a detailed account of various experiment... | 7 |
6mLjDwYte5 | [
"strength: This paper does not really present a new algorithm or theory, but is mostly a large collection of experiments showing under what conditions the proposed Flan-MoE model performs well. With that being said, the number and thoroughness of the experiments and ablations is quite impressive.",
"strength: The... | 6.75 |
z9UBpl4pv5 | [
"strength: Theoretical Foundation: The structured initialization method is based on solid theoretical analysis rather than just empirical results, providing a strong rationale for its effectiveness.",
"strength: Performance Improvements: The method consistently shows significant performance improvements over conv... | 4.333333 |
stUKwWBuBm | [
"strength: Writing is clear overall (I have some concerns for some aspects though, see next section) and the paper is well-organized. The proofs are well exposed in the appendix.",
"strength: Investigating the tractability of games combining bounded rationality and risk aversion is interesting. In particular I fi... | 8 |
HMe5CJv9dQ | [
"strength: The paper provides a simpler algorithm with a slightly different (better in some settings) trade-off for privately approximating distances and KDEs. The problem is an important one with many usecases. For this reason, I think the result is interesting",
"strength: The techniques used involve decomposin... | 7.5 |
L0pMPCmEfN | [
"strength: The result is interesting in very specific settings, such as e.g. the detection of Parkinson’s Disease and there is some novelty regarding the math. The fact that you go below the noise on T (i.e. the \"Weak Signal recovery bounds in Theorem 4.6.\") is interesting but it should be better explained and qu... | 4.333333 |
PFRWGeUhJx | [
"strength: The main text of the paper is well-written, and the main results of the submission are well-explained. The authors have put efforts to explain their technical contributions, in particular, the way how the estimates of normalized gradients of theorem 1 are propagated through the further results.",
"stre... | 4.75 |
fvse7bMkAs | [
"strength: The paper's topic of choice is timely: expanding on the available toolbox for quantifying and/or comparing the risks of multiple machine learning models, evaluated on multiple metrics is important in a context where foundation models are deployed in an unprecedented speed in various domains.",
"strengt... | 5.166667 |
NJ6nyv3XWH | [
"strength: The proposed method is simple and consistently outperforms baseline methods over multiple datasets.",
"strength: The authors provide a relatively clear and practical explanation for their approach.",
"strength: The authors provide ablations over different architectures for the GNN decoder.",
"stren... | 3 |
0iAZYF9hrl | [
"strength: The paper evaluates the recent ideas of disentangled representation learning using weak supervision in a more realistic application.",
"strength: The paper also presents an alternative to learning the disentangled representation from RGB images based on models pretrained at large scale.",
"strength: ... | 2.5 |
JePfAI8fah | [
"strength: This paper provided a simple and effective inverted view to improve transformer-based multivariate time-series forecasters.",
"strength: Compared with the previous use of Transformer structure (without invert), the iTransformer showed some advantages, including better generalization on unseen variates,... | 7.5 |
Mri9WIfxSm | [
"strength: The paper is well-written and clear.",
"strength: Developing efficient perplexity bound and the training algorithm for discrete diffusion language models is an important yet challenging problem.",
"strength: The particular forms of the bounds, the roulette diffusion matrix, the denoising cross entrop... | 6.75 |
Z91rwXnJsw | [
"strength: The paper proposes a complete architecture containing both learning-based and classic map-based planning methods to improve the performance on the visual object navigation task.",
"strength: The two approaches are combined using a decision module.",
"strength: Navigation is performed either by choosi... | 2 |
EukID7GvBy | [
"strength: Conducts several ablations on sequential multi-stage finetuning on easy to hard dataset subsets.",
"strength: Efforts in designing strategies for how we fine-tuning with the replay data and the training data. We can see that, some unknown data change to MaybeKnown, justifying the hypothesis.",
"stre... | 3 |
rcKzU0Vns0 | [
"strength: This paper proposes SISOM, the first approach designed to jointly solve OOD detection and AL.",
"strength: The experiments on AL and OOD detection are extensive.",
"strength: This study is significant in that it explores the practical applicability of deep learning by simultaneously considering both ... | 2.5 |
AoIKgHu9Si | [
"strength: Presents an innovative use of robustified ANNs to predict task difficulty and enhance images, aiding human perception and learning.",
"strength: Shows broad applicability by successfully testing across diverse domains, such as natural image classification, dermoscopy, and histology.",
"strength: Achi... | 6 |
fsW7wJGLBd | [
"strength: The use of an online game to collect human-generated adversarial examples for instruction-following LLMs is a novel and creative way to understand the weaknesses of these models.",
"strength: The creation of benchmarks for evaluating LLM resistance to prompt injection attacks is a valuable contribution... | 7 |
9Q9KXUTjmd | [
"strength: Comprehensive validation: In Sec. 6, FedTOGO is compared against a lot of FL algorithms and is able to outperform all of them.",
"strength: The main contribution of the paper is to propose a local and global perturbation-based algorithm (FedTOGA) that enjoys communication and computation efficiency for... | 4 |
PQrkWvQSL0 | [
"strength: Integration of heterogeneous data: The agent framework attempts to combine different information sources and prediction models as different agents, providing analysis from multiple perspectives.",
"strength: Practical application: The work includes case studies showing how the developed system can be a... | 2.5 |
todLTYB1I7 | [
"strength: This paper provides a detailed summary of existing work and unifies them under a mathematical framework.",
"strength: The experimental results of this paper reveal shortcomings in the majority of existing evaluation metrics.",
"strength: Unified Framework: The paper effectively consolidates various e... | 5 |
XHvguNJRbE | [
"strength: The focus on intrinsic motivations addresses a critical area in RL, allowing agents to develop behaviors aligned with real-world-like needs and goals.",
"strength: The paper adopts a first-principles approach, drawing insightful parallels between RL agents and humans by emphasizing the role of intrinsi... | 2.5 |
bntJK4NyIW | [
"strength: Leveraging underutilized, globally-distributed GPUs for training large-scale models looks promising and useful, especially for many individuals and organizations who cannot afford specialized clusters.",
"strength: The paper introduces a hybrid parallelism method with parameter servers to address the i... | 2 |
wZiH43e5Ah | [
"strength: The authors utilize various evaluation criteria to assess both their proposed method and existing methods, which adds robustness to their analysis.",
"strength: The paper is well-written and easy to understand.",
"strength: The paper provides extensive experiments to show the effectiveness of their p... | 3 |
BSqVfAFJWz | [
"strength: The paper addresses an interesting and important problem that is likely of interest to the community.",
"strength: The proposed method generalizes previous work in the area and provides a practical scheme for performing RL with a certain class of perturbed rewards.",
"strength: The method proposed fo... | 5 |
tjn2YZSHUv | [
"strength: Overall, this is an interesting paper on a topic which is of interest to ICLR Conference. It introduces an innovative set of metrics designed to gauge the intricate interplay between prompts and images, accompanied by the construction of an unprecedented million-user-scale dataset, offering invaluable in... | 6.75 |
U42TkrEDzb | [
"strength: This paper proposes to predict the MOS within the framework of a currently popular audio LLM scheme.",
"strength: The authors present their method clearly and effectively, detailing how the model assesses various aspects of audio quality before generating an overall MOS score.",
"strength: Experiment... | 6.75 |
H6i47PKXSN | [
"strength: Clear illustration of the model structure (Fig. 2), and abundant visualizations.",
"strength: The scaling experiments in Sec. 4.3 are intuitive and valuable, since scaling properties of autonomous driving models are previously underexplored.",
"strength: Outstanding performance of the proposed method... | 5.25 |
tZP4Uyql0r | [
"strength: The proposed Automatic Gradient Accumulation Module (AGAM) aims to regularize the classification layer through personalized gradient accumulation, enhancing the model's generalization capability on out-of-distribution samples.",
"strength: This module demonstrates the paper's effort to improve the mode... | 2.333333 |
mpqMVWgqjn | [
"strength: The idea of iteratively refining generated sequences from protein structures based on a confidence-aware gated fusion layer that fuses features from multi-modal feature extractors looks novel to me.",
"strength: Experiments across multiple datasets show the effectiveness of the proposed method.",
"st... | 6 |
94kQgWXojH | [
"strength: The paper introduces a novel application of the logit lens technique to interpret the internal image representations of VLMs, providing new insights into how these models process visual information.",
"strength: The proposed knowledge erasure algorithm, PROJECTAWAY, is a simple yet effective method tha... | 6 |
TbOcySs6g8 | [
"strength: The paper addresses a valid and important topic, and it’s generally well written",
"strength: The proposed method is straightforward to implement and readily applicable in practice.",
"strength: Experimental results are promising, highlighting the potential of using pre-trained models for DP data gen... | 2.5 |
Xq12wsoNux | [
"strength: The paper tackles an important problem of private training of large models.",
"strength: Fine-tuning large pre-trained models has proved to be an appealing avenue for mitigating the utility hit caused by DP training. The proposed pipeline extends the modern state-of-the-art DP fine tuning to very large... | 3.5 |
G536mmC2HL | [
"strength: Efficacy Demonstrated through Experimental Results: Empirical evidence substantiates TorSeq's exceptional efficiency when juxtaposed with other extant methods. This compelling empirical validation underscores the robustness of TorSeq as a pertinent solution within the domain of interest.",
"strength: I... | 3 |
IGzaH538fz | [
"strength: The theory of the paper is solid, and the experiments prove the point that the authors want to make.",
"strength: The presentation is pretty clear and easy for the reader to follow.",
"strength: Originality: The paper stands out for its innovative use of hashing to create subgraphs that distinguish b... | 8 |
tKFZ53nerQ | [
"strength: Contributes a prompt engineering approach to the task.",
"strength: Contributes an automated evaluation using BLEU and ROUGE",
"strength: Contributes additional experiments utilising video captions to see how performance varies.",
"strength: The quality of the research is commendable, as it employs... | 2 |
N1L5TgtkAw | [
"strength: The results in the paper are quite original and significant, to my knowledge, which is a bit limited.",
"strength: The experiments show that the proposed algorithm works well in practice.",
"strength: The paper is well written and easy to follow.",
"strength: Theorems 2 and 3 are novel and the expe... | 7.5 |
py3RTHNT6J | [
"strength: The paper is easy to read and the experiments studying scaling laws for multispectral images could be promising for remote sensing community.",
"strength: Also, the study shows that Swin Transformer based architectures are better suited for their tasks.",
"strength: Interesting problem setups (invest... | 2.2 |
rDLgnYLM5b | [
"strength: Propose a novel ISG framework and ISG-BENCH benchmark dataset that fill a notable gap in the evaluation of multimodal interleaved generation.",
"strength: The multi-level evaluation in ISG, combining visual question answering (VQA) with reasoning-based questions, offers a fine-grained approach to asses... | 7.2 |
VojvkUEq8q | [
"strength: Explicitly modeling propagation delay captures the lead-lag structure of causal relationships, allowing CALAS to more accurately reflect how signals influence each other over time.",
"strength: CALAS demonstrates efficient improvements over baseline models.",
"strength: One of the important factors i... | 4.75 |
fapHf9fmqp | [
"strength: The proposed method is fairly clearly well-explained.",
"strength: The empirical results on the anomaly detection benchmark are quite impressive and show the proposed method is useful in a number of real-world problems.",
"strength: Soundness: the paper establishes a sound theoretical framework for u... | 4.333333 |
NCrFA7dq8T | [
"strength: Clarity: The paper is clear and well-written, with high-quality illustrations and concise, precise language. The flow is logical, and the key points are emphasized effectively.",
"strength: Quality: The experiments are well-designed and effectively executed. The choice of mechanistic interpretability t... | 6.6 |
14E7S17hFv | [
"strength: In my view, the main strength of this work lies in the presented theoretical assessment. It shows that the minimum-action value leads to higher temporal difference (TD) than random actions and the difference in the TD is equal to the disadvantage gap. Such finding reveals the underlying importance of the... | 4.25 |
BjaHYhr7VS | [
"strength: Well-written, the intuition of this paper is clear, if the shape of a convex optimization surface is given, we can obtain the optimal learning rates for each directions directly. This can be done on top of any gradient momentum, and preconditioning regularizations.",
"strength: Although the dimension o... | 4.75 |
b57IG6N20B | [
"strength: It is a universal compression model for both iEEG and EEG.",
"strength: It achieves high-fidelity compression of EEG signals up to a compression ratio of 64.",
"strength: It has been validated both experimentally and being rated by the neurologists.",
"strength: The methodology is rigorously implem... | 6.6 |
C4QNyUqBIu | [
"strength: The paper is motivated by a real problem in flow networks, which is ubiquitous in communications, transportation and logistics. This potentially makes this work relevant to a wider audience, e.g., the communication networks community.",
"strength: The main observation behind PEW is meaningful. One expe... | 3.5 |
Xxpt66OgHI | [
"strength: The paper touches upon a very important challenge in training neural networks. To make training more efficient and scalable, there has been tremendous work on gradient compression to tackle the communication bottleneck in the distributed setup. However, over the past few years, some solutions have been p... | 4.333333 |
mYo9r0CwUf | [
"strength: It is nice to have a transmittance bound for each interval along the ray to facilitate the point sampling.",
"strength: The analysis of transmittance bound and divide-and-conquer sampling is interesting.",
"strength: The experiment results indeed show improvement over NeDDF.",
"strength: The paper ... | 2.333333 |
eQcVfCK5cO | [
"strength: Experiments are done on 3 datasets and show seemingly competitive results",
"strength: Adapt a diffusion model for trajectory generation looks interesting",
"strength: Some ablation studies are provided to support a few important design choices.",
"strength: The paper is well-organized and easy to ... | 5 |
9LAqIWi3QG | [
"strength: The paper is easy to read and describes the method succintly.",
"strength: The paper proposes a novel reward shaping method in the LLM setting.",
"strength: The experiment result is good.",
"strength: The idea of redistributing rewards at the token level in RLHF is compelling. Traditional RLHF meth... | 3 |
viC3cpWFTN | [
"strength: The paper is extremely well written. The position of this work to the related work is clear. The problem is clearly stated. There are very convincing examples, progressive explanations of the proof strategies, and the \"right level\" of details.",
"strength: Elementwise Gradient clipping is widespread ... | 5.333333 |
lnffMykYSj | [
"strength: The writing of the paper is clear, and original as far as I know.",
"strength: The main strength of the paper is the intuition it provides; the reasoning is easy to follow and well-motivated, and it sheds light on the performance gap between SSMs and transformers on long-range tasks, which is a very im... | 4.5 |
TqM0hifngW | [
"strength: The paper is novel and addresses real-world problems. More specifically, decomposing the behavior cloning into visual transition prediction and inverse dynamics learning is novel, prior approaches mostly rely on labeled data or direct behavior cloning, while JEPT makes use of unlabelled data and largely ... | 7 |
KaYXsoCxV7 | [
"strength: This article studies the Vision MoE architecture. The article is easy to read and has a clear structure.",
"strength: The shows that a simple method (adding sparse MoE layers, then fine-tuning) can improve the classification performance of pre-trained ViTs. It also presents a series of ablation study o... | 3 |
rGyi8NNqB0 | [
"strength: LiCE is a principled method of inferring *plausible* counterfactual explanations that satisfy several desiderata. The MIO approach is flexible enough to accommodate additional criteria.",
"strength: The MIO formulation for SPN inference is itself a significant contribution. It opens up space for work o... | 6.5 |
ZClm0YbcXP | [
"strength: Novel and well-motivated solution that specifically addresses MoE architecture challenges in unlearning, filling an important gap in the literature",
"strength: Thorough follow-up studies that justify key design choices, particularly: the effectiveness of single-expert versus multi-expert unlearning, a... | 5.25 |
5ncdKonxd4 | [
"strength: The proposed method improve LVLM efficiency through progressive token reduction based on layer-wise redundancy analysis. Different from compressing tokens uniformly or only in early layers, PyramidDrop's stage-wise reduction aligns with the redundency distribution of visual tokens in LVLMs.",
"strength... | 3 |
NzEIjnIIzv | [
"strength: The proposed method is simple but effective.",
"strength: Comprehensive ablation results with reasonable baselines.",
"strength: The paper includes a lot of ablations that are very convincing of the efficacy of their approach. As I was reading, I keep thinking, I wonder if the see effect is because o... | 7.25 |
CdqQKXGKq3 | [
"strength: The structure of the paper is logical, with clear main text and conclusions, making it easy for readers to understand and follow.",
"strength: The diagrams in the paper are aesthetically pleasing and easy to read.",
"strength: The paper is clearly written and motivates the proposed approach well in a... | 4 |
IzYczpPqKq | [
"strength: The method introduced in the paper is commended for its intuitive and logical approach. By assuming that the learning dynamics of the agents are Markovian and can be learned in finite steps, the paper effectively simplifies the complex problem of steering under uncertainty.",
"strength: The paper provi... | 6.333333 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.