paper_id stringlengths 10 10 | items listlengths 4 82 | gt float64 1 9.5 |
|---|---|---|
Pj06mxCXPl | [
"strength: The paper argues with proofs and empirical evidences for its results and is able to remove any possible doubt of their veracity. Furthermore it's a novel exploration of the capabilities of linear transformers and further helps the understanding of In-Context Reinforcement Learning.",
"strength: The aut... | 6.666667 |
Hh0Cg4epYY | [
"strength: The problem of upper bounding Bayes error is somewhat interesting, though I am not sure how important it is. The paper does not present sufficient evidence in handling it.",
"strength: The overall approach seems like an interesting idea, as it suggests a new objective for optimizing complex classifiers... | 2.333333 |
rMR2P8e0Zx | [
"strength: The paper introduces both a novel model, SongComposer, and a comprehensive dataset, SongCompose, which are substantial contributions to the field of AI-driven music composition. The dataset includes a large collection of lyrics, melodies, and precisely aligned lyric-melody pairs. The data would greatly b... | 4.6 |
LC2KxRwC3n | [
"strength: This paper studies an interesting problem, that is, learned sparse features can become less understandable by \"absorbing\" token-aligned features.",
"strength: understanding how width and sparsity of SAEs affects its training performance can be helpful for future SAE training.",
"strength: the paper... | 7.5 |
eRAXvtP0gA | [
"strength: Proposal of a novel representation-centric unsupervised learning algorithm: The paper introduces a new approach that focuses on generating constructive representations of the input space. This algorithm transforms inputs into Sparse Distributed Representations (SDRs) and organizes them hierarchically.",
... | 2.5 |
07yvxWDSla | [
"strength: The problem the work addresses is important.",
"strength: Experimental results show that this method scales better than simple paraphrasing or direct pretraining, and that retrieval-augmented generation further boosts performance of this model.",
"strength: The authors also present a theoretical mode... | 8 |
rTBL8OhdhH | [
"strength: The proposed framework is quite well motivated, and the experiment results aligns well with the observation",
"strength: In the experiments, the authors are the first to demonstrate that distilled synthetic data can outperform real training dataset",
"strength: The authors perform extensive studies a... | 7 |
K3n5jPkrU6 | [
"strength: The paper introduces a pioneering framework that employs directed acyclic graphs (DAGs) to organize agents within a multi-agent system. It provides insightful comparisons between various topologies.",
"strength: The research delves into the impact of scaling the number of collaborative agents on overal... | 7 |
bAFVlpFQvT | [
"strength: The approach takes a unique and innovative perspective on memory-efficient training. Unlike typical methods in field, it focuses on data selection and reduces batch size to achieve memory-efficient training.",
"strength: The paper is clear to understand and well motivated.",
"strength: The paper pres... | 6.75 |
pNdPJACSLB | [
"strength: Clear time series problem definition in the preliminaries",
"strength: Interesting method to use modfiy the gradient updates and study the optimization method in U-Nets",
"strength: The introduction of transformer based models has significantly increased the sizes of SOTA models for time series forec... | 2.5 |
4YzVF9isgD | [
"strength: Well-formulated optimization problem: The paper effectively defines the optimization problem for generating high-quality synthetic face recognition datasets.",
"strength: Efficient solution: The proposed solution is not only effective but also computationally efficient.",
"strength: Extensive experim... | 5.25 |
fcSDt7H8kI | [
"strength: This paper proposes an interesting idea of improving the exploration efficiency by taking extremum action, which refers to the action with minimum Q-value.",
"strength: The method comes with a nice theoretical motivation, where the authors show the proof of the relationship between TD error inferred by... | 4 |
DcG4YnbOT3 | [
"strength: The paper introduces a novel approach to time series forecasting by framing the problem within the image modality, offering a fresh perspective.",
"strength: The model structure is clear and appears logically sound.",
"strength: The presentation is well-organized and easy to follow.",
"strength: Co... | 4.333333 |
EaB7Ue1X9p | [
"strength: The paper aims to address a very important problem.",
"strength: Existing literature has also been sufficiently reviewed.",
"strength: Most importantly, I really appreciate that the paper features a pretty interesting set of very real-world, practical experiments.",
"strength: There is also a resul... | 5.25 |
bXI0thP733 | [
"strength: This paper builds on an empirical observation that comparing per-class densities over extracted features can reveal poisoning behaviour.",
"strength: The ablation study shows that choice of feature representation doesn’t significantly vary the performance with two models resulting in marginal differenc... | 4.5 |
4vzGQcVUG8 | [
"strength: The math appears correct to me; the problem is significant, and desiderata 1 and desiderata 2 make sense.",
"strength: Exact characterization of the regime where weak to strong generalization occurs in terms of parameters of the covariance matrix of strong and weak features.",
"strength: This work ad... | 6.5 |
jX2DT7qDam | [
"strength: Analysing the conformal intervals for the early exiting classifiers is interesting and it's promising to see the proposed method yields better uncertainty characterisation.",
"strength: The ablation study in 8.6 clearly shows the merits of learnable GMs versus other design choices.",
"strength: Figur... | 7.5 |
84fOBZlOiV | [
"strength: The paper addresses a critical problem in the field of machine learning – the propagation and quantification of output uncertainty through neural networks based on input uncertainty.",
"strength: This is an increasingly important issue in modern machine learning, as it directly relates to the reliabili... | 4 |
WcOohbsF4H | [
"strength: The paper tackles an important problem since high quality ECG labeled data are scarce but ECG data in general is much more available.",
"strength: The proposed solution is simple yet very effective as shown in the results, also compared to other SSL alternatives.",
"strength: The paper was overall cl... | 7 |
zQXX3ZV2HE | [
"strength: A novel adversarial attack is defined for human object interaction recognition task. HOI recognition task is a high-level task with respect to object classification or detection tasks. It motivation of this paper makes sense which is to insert an inviable object to mislead the HOI classifier.",
"streng... | 3 |
XOnya9gSdF | [
"strength: The proposal is theoretical sound, and detailed analysis of the properties of proposed metrics as well as how to build the consistent classifier based on confusion matrix measure and tensor representation, are shown. One practical solution for optimising the classifiers with proposed macro metrics, is de... | 7.5 |
PBjCTeDL6o | [
"strength: Authors well-summarize the previous search, and clearly explain the required properties needed for the reference. The intution is clearly written, and well-evidenced by experimental results.",
"strength: The paper offers a well-structured analysis of an important problem in neural network interpretabil... | 8 |
uw5U7FfTRf | [
"strength: Addressing the datasets backdoor attack is still an interesting and realistic direction.",
"strength: Detecting multi-trigger backdoor attacks is an efficient way to deal with large scale dataset.",
"strength: The use of loss trajectory to detect backdoor samples is an interesting direction;",
"str... | 3 |
5e0yWSNGIc | [
"strength: The paper is well-organized and easy to follow.",
"strength: The research problem is interesting, it aims to answer the question \"What cost does adversarial training bring in value estimation of DRL? \"",
"strength: The motivating linear model example well explains the angle of overestimation and wr... | 5.333333 |
cMLtjP3Cym | [
"strength: This work improves upon previous guidelines by prioritizing both channel-independent and diffusion-based generative frameworks, providing users with a more comprehensive and practical TSGGuide for selecting time series generation frameworks.",
"strength: By enhancing the central discriminator in the ch... | 4.75 |
4ymHtDAlBv | [
"strength: Easy to follow and clearly written;",
"strength: Lightweight solution for text classification compared to transformer models.",
"strength: This paper proposes a structure that is 5 times faster to train than LSTM.",
"strength: The paper is well-organized, and the writing is fine.",
"strength: The... | 2.333333 |
U42IGDU3q5 | [
"strength: The paper is timely for the CIR community, I found the paper well structured and well written.",
"strength: CIRQRS addresses a core limitation in current CIR evaluation metrics by focusing on user-perceived relevance rather than simply on retrieving a target image. Traditional metrics like Recall@k fai... | 5.333333 |
feykAFeSfq | [
"strength: RL algorithm for finding the optimal keyboard layout is a new angle to approach the problem.",
"strength: In the fields of Human Computer Interaction, the paper holds great potential where a lower cost layout will not only improve user experience, but also improve productivity. From that perspective, t... | 3 |
0fJfVOSUra | [
"strength: The paper proposes methods at different levels that simplify CUDA kernel implementations",
"strength: The paper can achieve a similar performance compared to the state-of-the-art implementation",
"strength: The authors demonstrate significant improvements to computational efficiency within a clearly ... | 7.5 |
Nifg2fQMGW | [
"strength: The method can be applied to various Text-to-Video models, clearly stating the intended goal of simultaneously editing the content's motion and appearance.",
"strength: The paper presents results applied to multiple Text-to-Video models. Experiments using various prompts demonstrate that this method is... | 4.75 |
WOzffPgVjF | [
"strength: The intuitive motivation of TA-STVG is clear and easy to understand, i.e., the adaptively generated object queries containing target-relevant information facilitate the spatio-temporal localization of textual expressions.",
"strength: The proposed method to enhance and keep track of the information abo... | 7.5 |
UzgMX1rwGc | [
"strength: The paper introduces a novel distribution-based placement network, going beyond simple heuristic placements, resulting in highly plausible object positioning.",
"strength: The method is evaluated on both KITTI and NuScenes datasets, showcasing clear performance gains for monocular 3D detection.",
"st... | 3.5 |
DwiwOcK1B7 | [
"strength: The idea is nice, the problem formulation is neat, and using ADMM for optimization is elegant.",
"strength: On pruning LLAMA, the method shows clear benefit over the compared methods. Image classification experiments are marginally better than previous methods.",
"strength: While pruning of factors o... | 6.333333 |
xvUVk9T3kZ | [
"strength: The problem of recovering reward functions using demonstration is interesting as this will facilitate better transfer and generalization across tasks.",
"strength: The experiments on Meta-World show that the method can recover the reward functions that are common across tasks.",
"strength: **[Motivat... | 3 |
fiTpna7fO5 | [
"strength: Novelty: The paper introduces a compelling and innovative approach by applying extensive constraints to the adversarial problem. This perspective challenges the common belief that adversarial examples are rare and unique, demonstrating instead that a multitude of such examples exists across various regio... | 3.25 |
2JXe3RprGS | [
"strength: This paper introduces a deep learning model for turn-by-turn navigation, offering two key benefits: more adaptive, context-aware audio guidance and reduced navigation errors. The model’s sequence-based design and cloud-edge setup ensure real-time, precise directions, greatly enhancing driver support.",
... | 3 |
PXNrncg2DF | [
"strength: Unsupervised image segmentation holds significant importance, and this study successfully performs segmentation without human supervision, offering segmentation masks at multiple levels of granularity.",
"strength: The approach to generate unsupervised hierarchical masks is pretty interesting. And surp... | 6 |
hPpyUv1XyQ | [
"strength: The paper has a fairly extensive number of experiments.",
"strength: This work demonstrate the use of MBR considering the uncertainty in model parameters. The idea is straightforward and the proposed approach of two decision rules is a sound method in performing either summation across samples drawn fr... | 5.25 |
5BXAXOpaWu | [
"strength: This paper is of good written quality that makes the readers easy to follow. The logic, the notion expression, and the experiments are all very clear.",
"strength: The asymmetric design is interesting and this design has been proven an efficient way to deal with resource-limited circumstances.",
"str... | 7.5 |
gbruScKTJ2 | [
"strength: This work addresses an interesting and important question about how global scenes vs. objects are represented in visual processing and whether evidence from neuroscience/cognitive science can help design better deep learning architectures.",
"strength: Specifically I find the idea of implementing a dua... | 3.666667 |
I0mQlersGk | [
"strength: Through experiments and competitive results",
"strength: Incorporating graph Transformer and spiking neural networks through spiking vector quantization",
"strength: This paper proposes a new spiking vector quantization for efficient graph transformers.",
"strength: The proposed model achieves comp... | 4.75 |
v1B4aet9ct | [
"strength: To my knowledge, the proposed method is original. I like the proposed SPODNET, building on classical theory, simple but elegant.",
"strength: The paper is well written and well structured.",
"strength: Experiment design is appropriate for demonstrating the effectiveness of the proposed method. I part... | 6.666667 |
Z1E0EahS5w | [
"strength: Reservoir computing is an efficient approach for learning dynamical systems, and the work aims to provide more insights into this class of methods.",
"strength: The paper framework looks solid. The approach of breaking the sequential nature of reservoir computing to classification is a good reduction m... | 3.333333 |
mGHJAyR8w0 | [
"strength: The problem studied in this paper is topical and practically meaningful, as geometric GNNs are quite widespread in application.",
"strength: The insights presented in the paper are prescriptive, and give a rule of thumb for selecting hyper parameters such as the highest “L” (indexing the irrupts of O(3... | 6.5 |
4A5D1nsdtj | [
"strength: [Novelty] The main idea of designing homophilous-related bases is insightful and instructive. By investigating the correlation between homophily and bases, the well-designed heterophilous bases can adaptively and effectively address heterophilous graphs according to the homophilous ratio.",
"strength: ... | 4.5 |
XmProj9cPs | [
"strength: Originality: Spide 12.0 significantly extends traditional text-to-SQL benchmarks by focusing on real-world enterprise scenarios. It incorporates a broader and more complex range of challenges, such as handling extremely large databases, diverse SQL dialects, and intricate workflows. This originality lie... | 8 |
g3PuaFh5vV | [
"strength: The promising study introduces a large body of experimental effort on a relatively less explored field, MEG data.",
"strength: The code and pipelines in this work can contribute to increasing interest in the field.",
"strength: Source space reconstruction of MEG data is a solid goal, albeit a well-st... | 2.5 |
324fOKW1wO | [
"strength: The paper uses decision transformer, with a set of practices in online RL (prioritized replay buffer), to address the closed-loop planning task in Waymax.",
"strength: The introduction of multi-token prediction in a decision transformer framework is interesting and may help with the realtimeness of the... | 3.333333 |
4u0ruVk749 | [
"strength: Paper is easy to follow, formulations and assumptions are clearly layed out.",
"strength: Algorithmic workflows are presented in detail.",
"strength: Experiments are done thoroughly with various benchmarks, and visualization of the results helped understanding model performance.",
"strength: Overal... | 3 |
PdaPky8MUn | [
"strength: The paper does a fairly through evaluation of SPT and its impact on evaluation on various benchmarks. The evaluation clearly shows gaps in the current evaluation of different architectures on long sequence tasks.",
"strength: In addition to providing guidance on evaluation practices the experiments als... | 8 |
V1N6MmDY27 | [
"strength: The paper highlights alternative approaches to address the challenges of out-of-distribution data in machine learning models.",
"strength: It conducts tests and analyses on standard evaluation settings and provides source code for the experiments as supplementary material.",
"strength: The paper pres... | 2.5 |
473sH8qki8 | [
"strength: The paper is well written and easy to understand",
"strength: The method is principled and intuitive.",
"strength: The authors perform ablations and provide statistically significant results.",
"strength: The paper is very well-written and very easy to follow and understand.",
"strength: Thorough... | 2 |
natXOadi7j | [
"strength: The research question is relevant since communication is a frequent limitation of FL protocols.",
"strength: The research question is well-motivated by Section 3.2.",
"strength: The analysis of sample-level DP with several local steps seems novel.",
"strength: The analysis and proofs are easy to un... | 4.666667 |
kkVTeMvC9D | [
"strength: Understanding the training dynamics of deep neural networks is interesting and can have lots of potential in accelerating both training and inference of these large models.",
"strength: The paper is well written and most of the figures are clear.",
"strength: The paper is well written.",
"strength:... | 3.4 |
GOjr2Ms5ID | [
"strength: A cascading (learned) Bloom filter (LBF) architecture that combines both sandwiched LBFs and partitioned LBFs (PLBFs), with classical BFs (as branch BFs).",
"strength: The article proposes a novel structure of CLBF that optimizes memory usage and query performance by cascading multiple machine learning... | 3.25 |
xQCXInDq0m | [
"strength: The paper is well written and easy to read.",
"strength: The proposed approach to achieve personalization is simple, novel, and training-free, applicable to various LLMs.",
"strength: Extensive experiments demonstrate strong performance in personalized recommendations, identification of implicit inte... | 6.666667 |
BW8O4wHgbo | [
"strength: The paper is well-structured, and it clearly explains why LLMs may be suitable candidates for MAPF due to their reasoning capabilities and large contextual understanding.",
"strength: It conducts experiments on various MAPF benchmark maps, evaluating LLMs’ performance in different scenarios and identif... | 3 |
Aku2I3z4aV | [
"strength: The IFGW distance represents a significant advancement in similarity metrics for graph data, integrating both structural and feature information in a single optimal transport framework.",
"strength: This metric has the potential to inspire further research into improved similarity measures for structur... | 2.6 |
PRJ4n3CBzU | [
"strength: Build high-quality, reliable dataset AttackQA for cybersecurity community",
"strength: Plans for dataset and benchmark open-source",
"strength: In addition to the dataset generation, a significant amount of work has been done to illustrate feasible application scenarios for AttackQA, enriching the hi... | 4.25 |
nUBLhhVM1l | [
"strength: The outlier detection in transfer learning is an interesting and valuable topic in the learning community.",
"strength: The literature part is very clear.",
"strength: The structure of the paper is easy to follow.",
"strength: The setup of the paper is clear.",
"strength: The paper provided solid... | 6.333333 |
RmQAKu1wCe | [
"strength: The MTT performs a wider search along the timestep setting by sampling different.",
"strength: Compared to naive NMT, MTT urges each layer of SNNs to learn feature that is weakly correlated to input length from the previous layer.",
"strength: Such a strong regularization causes an issue in assessing... | 5 |
OcTUquFXfx | [
"strength: The authors present an new approach by utilizing the action functional to reformulate the loss landscape into a more manageable shape. This idea of transforming a \"rough\" loss landscape into a \"benign\" one for optimizing neural network parameters is neat and innovative and could offer many advantages... | 2.6 |
XsYJ6yvgEC | [
"strength: The paper proposes unconditional and conditional evaluations, allowing for a nuanced comparison of generative models. It accounts for model drift and evaluates market impact metrics, making the framework robust.",
"strength: The use of widely adopted LOBSTER data for empirical evaluations makes the ben... | 3.333333 |
QCY1WQXTc8 | [
"strength: The proposed loss is simple and easy to implement.",
"strength: The paper also provides proofs to show SimO loss is a semi-metric and can help mitigate dimensionality collapse.",
"strength: Paper shows a number of t-SNE figures to showcase the learned embeddings.",
"strength: The paper aims to tack... | 3 |
SgymXhOEA5 | [
"strength: A thorough investigation of camera bias in reID is a big contribution. This paper provides a quantitative view for bias measuring, which is most welcomed.",
"strength: The consequently proposed debiasing normalization is very simple yet effective, with sufficient experimental and theorical analysis.",
... | 7.5 |
6E0x0lVvh8 | [
"strength: I think the paper is easy to follow, reasonably well structured, and mostly well written.",
"strength: The paper seems to include a pretty comprehensive discussion of related work.",
"strength: This paper studies five research questions related to probing language models in the context of theory of m... | 4 |
RiQRUcjXBD | [
"strength: The paper is generally well-written and easy to follow. The method and experiment sections are clear.",
"strength: The paper creates a new dataset with 50k AI/ML papers. I think it is a nice touch to annotate each paper and include entities such as main ideas, backgrounds. To my knowledge, no comprehen... | 3.5 |
i2r7LDjba3 | [
"strength: The paper contains novel contribution in that it is the first paper to present Echo video generation conditioned on ECG signals, and also plan to release the ECHO videos paired with ECG dataset.",
"strength: The paper is well-written and easy to follow, with the pipeline figure covering all three steps... | 6.8 |
HZtBP6DZah | [
"strength: Paper is well organized and written.",
"strength: Good experimental results on four datasets.",
"strength: Sound technical backbone.",
"strength: This paper adds new components to the works of Chen et al. (2022); Li et al. (2022a;b), and I particularly found the application of contrastive learning ... | 3 |
BUpdp5gETF | [
"strength: Has a valid and strong motivation.",
"strength: The finding that the RLRS's of a smaller model can be transfered (with slight adjustment) to a larger model without losing much fitness on the larger model is intriguing and practically meaningful.",
"strength: The experiments are comprehensive and do s... | 2.5 |
LzycEbgLoi | [
"strength: The paper presents a novel CTViT-V encoder, specifically optimized for 3D CT images, addressing the challenges of spatial coherence and slice dependency, which could contribute to improved clinical relevance and accuracy",
"strength: The use of GPT-4 for preference optimization provides a scalable, co... | 5.25 |
lvw3UgeVxS | [
"strength: The methodology is evaluated using both in silico and wet lab approaches. For in silico evaluation, three metrics (sequence recovery, self consistency, and perplexity) are reported. These metrics are described on page 16. The “wet lab” experiments are run on Eterna, an independent online computational RN... | 7.5 |
pcIDLhnYL9 | [
"strength: This work proposes a novel approach for model selection by relying solely on training loss to predict out-of-distribution generalization. This is different from the traditional validation-based model selection and is particularly suitable for scenarios involving distribution shifts.",
"strength: This p... | 3.75 |
x9cXrOQskc | [
"strength: The dataset preparation (collection followed by data cleaning) is at the centre of contribution.",
"strength: A new benchmark dataset is constructed specifically for station-based weather forecasting tasks.",
"strength: The process of dataset construction is clearly described.",
"strength: Experime... | 4.25 |
HE9eUQlAvo | [
"strength: Overall, this paper provides a comprehensive empirical demonstration of how to improve a performance metric of interest given training samples and model parameters.",
"strength: The breadth of properties: This paper considers several interesting scenarios ranging from noisy labels, active learning, adv... | 6.4 |
12iSWNLDzj | [
"strength: This paper addresses an important issue, namely the security of face recognition models, which is meaningful for both academia and industry.",
"strength: The paper leverages the ability of diffusion models to generate realistic and natural images to create adversarial face masks, which is an meaningful... | 3 |
Fb93MfxX7T | [
"strength: The paper addresses important questions in the field of PETL, contributing meaningfully to the ongoing discourse.",
"strength: The finding that weight-space ensembles (WiSE) could enhance performance on the target distribution is intriguing and merits further investigation.",
"strength: The writing o... | 4.75 |
ZZwP9zljas | [
"strength: The paper is well presented and well written.",
"strength: There is a good amount of literature review and context provided as well.",
"strength: The authors study different scenarios and overall provide a pretty detailed mathematical understanding of the topic they wish to study.",
"strength: This... | 5.25 |
iX1RjVQODj | [
"strength: learning a reward model from human preferences has flaws. And on top of that, using RL to optimize for this reward model can sometimes lead to poor performance. This paper solves this issue by not having a reward model.",
"strength: CPL has supervised objectives so it is scalable",
"strength: The pro... | 7 |
yarUvgEXq3 | [
"strength: Well-written and organized.",
"strength: Solid derivation and equations.",
"strength: Going beyond standard ERM objectives, nice mathematical re-formulation of tail risk CVaR and combining it with Matrix Factorization for Recommender Systems.",
"strength: The authors propose a novel and efficient a... | 7.333333 |
l4fMj4Vnly | [
"strength: Introduction of a new task - The paper addresses a previously underexplored area by defining the audio exploration task and the paper motivates this task well.",
"strength: Dataset creation and availability - The authors took great care in creating new datasets ACD and CLD for this task - which is a s... | 7.5 |
ZyAwBqJ9aP | [
"strength: The intuition of exploring sequence models and graph models for the task is rational.",
"strength: The novelty of this work is in combining two powerful neural network encoders for enzyme substrate prediction.",
"strength: A five-fold cross-validation is performed to select the best set of hyperparam... | 2 |
W4q7cwRCwg | [
"strength: This paper is very clear and easy to understand.",
"strength: The experimental section includes many baseline methods and datasets.",
"strength: Improving virtual nodes using a Gram matrix is an interesting idea, and the proposed model appears to perform well on benchmarks.",
"strength: The paper i... | 3 |
4QtywskEyY | [
"strength: The paper is well written in most parts.",
"strength: The proposed method is simple and straightforward.",
"strength: Comparative experiments are conducted on image classification (with CIFAR-100 and ImageNet datasets) and object detection (with PASCAL VOC dataset) tasks.",
"strength: The proposed ... | 6 |
kKXIYUi8ff | [
"strength: Originality: Combination of DDPM and molecular dynamics. The original part of DDPM is to introduce two variants of sampling. Moreover, a physics-inform layer is introduced to the Unet block.",
"strength: Quality: A lot of essential parts are missing.",
"strength: Clarity: The writing is poor. Many ty... | 3 |
5AJ8R4z5g0 | [
"strength: The issue is important, and the setting where covariate information is not available for the RCT does appear in unfortunately many applications.",
"strength: this paper studied an important and practical problem of CATE estimation, where the estimator can be significantly biased due to the effects of u... | 3.25 |
JVFRwCx3Dy | [
"strength: Timely topic and a relatively straightforward methodology for studying it.",
"strength: I liked Figure 1 with plots of n-gram repetition in real corpora -- it's a good motivation for the toy setup in the paper. (I was keen to see a deeper investigation along those lines but only found toy experiments.)... | 4 |
07xuZw59uB | [
"strength: The authors propose a group fairness and in-group individual fairness problem.",
"strength: The written of the paper is clear.",
"strength: The authors design clear experiments to demonstrate the advantage of the proposed method.",
"strength: The proposed framework FairGI can optimize the group fai... | 3 |
sq5LLWk5SN | [
"strength: This paper proposes a new method for distributional robust optimization.",
"strength: The mathematical framework is well-developed with thorough theoretical analysis. This paper provides two sets of theoretical results.",
"strength: The first one is the generalization/robustness bound. The main idea ... | 6.5 |
0qexTTfnmH | [
"strength: The proposed method is straight forward and the computational savings, compared to BLoB, are immediately obvious.",
"strength: It is also commendable that the authors have undertaken the task of reproducing the methods and experiments from the original BLoB paper.",
"strength: The core idea of this p... | 3.75 |
cUFIil6hEG | [
"strength: I think the construction of the graph whose isomorphic permutations of nodes preserves the model functionality, for transformer layer is neat and interesting.",
"strength: The paper is sufficiently well written and is fairly accessible.",
"strength: The proposed approach is sufficiently sound and no... | 5.75 |
pWrcpPsVas | [
"strength: The paper is very well-written, and all network design decisions are motivated out of the demands by the physical problem which are introduced in the requisite detailed manner.",
"strength: Especially the description of the physical problem is very well-written and enables an easy understanding of the ... | 4.25 |
XTBdPLhiRL | [
"strength: The paper is well-written and well-organized.",
"strength: The proposed method shows better results in various tasks, compared to previous methods.",
"strength: Clear summary of existing key challenges for event camera data pretraining.",
"strength: Clear writing, logic flow, and good content arran... | 4.25 |
tTPHgb0EtV | [
"strength: The research topic on tackling harmful fine-tuning is important and timely because of the urgent need to ensure the trained LLM can resist alignment attacks.",
"strength: The proposed approach is intuitive and clear, based on the clear definition of harmful perturbation. The experimental results improv... | 8 |
vePZdNvrO9 | [
"strength: The paper is articulate and well-organized, with clear definitions of key concepts and a logical flow of ideas. The use of the Chameleon Game as a case study helps in concretely demonstrating the application of GAMEINSTURCT, making the complex concepts more accessible to the reader.",
"strength: GAMEIN... | 4.25 |
1qbZekXGrp | [
"strength: The paper explores the potential of applying diffusion models to the REG task, an area that has been largely underexplored.",
"strength: The authors introduce a vision-text conditioning module and a token selection strategy, which significantly enhance the alignment between visual and textual informati... | 6.5 |
hYEV8QmaOt | [
"strength: The paper targets an important problem of image antiforensics where the goal is to hide traces such that it will be difficult for image forensics algorithms to reliably detect image forgeries. The problem itself is important and is usually less addressed in academic studies.",
"strength: The two-phase ... | 3.4 |
IwPXYk6BV9 | [
"strength: The paper proposes a simple method and shows that it outperforms prior methods proposed in literature of LP-MST (Label Privacy Multi Stage Training [[Ghazi et al. '21](https://arxiv.org/abs/2102.06062)]) and ALIBI [[Malek et al. '21](https://arxiv.org/abs/2106.03408)].",
"strength: The method is also s... | 7.5 |
di52zR8xgf | [
"strength: SDXL demonstrates impressive text-to-image generation results. It can serve as a strong base model for a broad range of research and applications in downstream image synthesis tasks.",
"strength: Some training and architectural choices are backed with convincing ablation experiments (e.g. Table 2). Thi... | 8 |
5RUM1aIdok | [
"strength: The proposed approach of evaluating the research ideas over the graph-structure (by breaking down them into multiple sub-ideas connected with other sub-ideas) is novel and sound.",
"strength: The proposed approach is superior to other research idea evaluation methods.",
"strength: This paper is very ... | 6.75 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.