paper_id stringlengths 10 10 | items listlengths 4 82 | gt float64 1 9.5 |
|---|---|---|
B2ChNpcEzZ | [
"strength: The work presents a simple and straightforward method that changes the prompts to a more carefully designed prompt using taxonomies.",
"strength: Fine-grained analysis and investigations are conducted on the different factors of building good semantic prompts, which can offer readers with some practica... | 4 |
RacYdzHxcz | [
"strength: This paper delves into a captivating avenue for generating adversarial perturbations.",
"strength: Moreover, the authors have thoughtfully crafted a robust loss function aimed at ensuring that the adversarial lines are feasibly replicable by humans.",
"strength: The writing quality is high and only a... | 3.5 |
XnDyddPcBT | [
"strength: The paper makes an interesting attempt to bridge the gap between theoretical analysis of transformers and practical implementations by using neural ODEs with depth-varying weights.",
"strength: The proposed adaptive fine-tuning approach, which leverages the discussed ODE formulation to enable flexible ... | 5.75 |
F4meTCwlxZ | [
"strength: The main strength of the paper is the use of causal constraints to evaluate putative causal relationships extracted from text. Prior work has focused mostly on acyclicity constraints, I've never seen a method take constraints to this level of rigor.",
"strength: For example, the LLM prompting strategy ... | 5 |
GwJXJSCH1S | [
"strength: This paper is interesting in researching on the size of edited subject to make it harmonious with original video.",
"strength: Using pretrained models to get world knowledge of subject size is reasonable.",
"strength: The proposed method achieves much better performance than previous methods.",
"st... | 3.75 |
X6W5eqhzDx | [
"strength: The paper proposes a creative solution to generalization in RL by leveraging exploration to maximize reachability of tasks. The exploration phase before each episode adds a distinct approach to state exploration, directly improving sample-efficiency (test performance on reachable states) and zero-shot in... | 4.666667 |
8OcM1pTfHm | [
"strength: By eliminating reliance on costly fine-tuning and human-annotated preference data, AlignEZ offers a more resource-efficient alternative to traditional alignment techniques, making it potentially scalable for larger or real-time applications.",
"strength: AlignEZ presents a conceptually straightforward ... | 5.333333 |
ERBm5WK8nq | [
"strength: This paper offers a new perspective for LLM-enhanced time-series forecasting models. Unlike existing methods that aim to align LLMs directly with time-series models, the proposed method integrates multimodal information to improve the ensemble of multiple linear experts.",
"strength: Experiments demons... | 3.5 |
IqaQZ1Jdky | [
"strength: The paper introduces the Variety Bernstein Polynomial Function Basis for Kolmogorov-Arnold Networks (VBn-KAN), which is a novel approach in the context of KANs.",
"strength: The authors provide a solid theoretical foundation based on the Weierstrass approximation theorem and propose that Bernstein poly... | 2.5 |
WNzy9bRDvG | [
"strength: This paper is really well-written and easy to follow.",
"strength: Each component of the design space is carefully explained and well-presented.",
"strength: Most of the choices are accompanied by intuitive demonstrations, which provide insights that could translate to other methods.",
"strength: M... | 7 |
DugT77rRhW | [
"strength: The method is sound with great results outperforming other methods.",
"strength: This is the first method for unposed sparse view room layout reconstruction, especially when the views are not overlapping.",
"strength: The whole idea is interesting, especially to use structural plane depth map and met... | 6.75 |
i3e92uSZCp | [
"strength: A theoretical proof of language-distance as a valid pseudometric is provided.",
"strength: Both locomotion and manipulation tasks have been demonstrated, showing the effectiveness of the proposed approach.",
"strength: The paper is well presented.",
"strength: The method tried to link skill states ... | 7 |
GCH5leffZp | [
"strength: Originality: The NeRT framework models periodic signals by effectively separating periodic and scale factors. This method creatively combines concepts from Fourier analysis and neural network architectures, leading to a more nuanced understanding of periodic behaviors in data.",
"strength: Clarity: The... | 7 |
mJ8k81O5BF | [
"strength: The synthetic image generation process without BN statistics supervision is helpful for networks without BN, also simplifying the optimization process.",
"strength: This paper uses very few samples, no more than 35.",
"strength: The paper is easy to follow.",
"strength: The proposed method can be e... | 3 |
0sJ8TqOLGS | [
"strength: The work uses inspiration from cognitive science to come up with framework",
"strength: They address an important aspect of LLMs which is critical thinking",
"strength: Multiple models are considered for the work and comparison",
"strength: Multiple hypothesis are tested in this work.",
"strength... | 5.25 |
GOt2kP383R | [
"strength: The motivation is clear and strong.",
"strength: The authors design a new loss term and channel-wise quantization factors to regularize the activation to make the network easy to quantize.",
"strength: The plug-in module can be introduced and gain improvements in other networks and tasks that has var... | 5.25 |
I1484gDBr4 | [
"strength: Finding alternatives to transformers is important, and LRNN and S4 are promising approaches.",
"strength: Clear explanation of the FST concept",
"strength: Overall, the paper is clearly written.",
"weakness: The paper seems to be rushed; the writing needs improvement.",
"weakness: Incremental wor... | 2.5 |
58KF6ne6d4 | [
"strength: Interesting idea of using RL to solve CNC routing problem which is traditionally solved by hand crafted algorithms",
"strength: The problem this paper aims to solve, i.e., integrated toolpath smoothing and federate planning, is rooted in real-world manufacturing. It is important for improving machining... | 3 |
GLuzjuG0lo | [
"strength: I don't see any obvious advantages compared to other latest medical image segmentation methods.",
"strength: The manuscript is well-organized and written",
"strength: Very detailed literature reviews",
"strength: Innovative Non-Causal Modeling: Unlike traditional state-space models that are limited... | 3 |
KZgo2YQbhc | [
"strength: This paper is well-structured and easy to understand.",
"strength: The idea of achieving T2I model personalization by controlling the rank of diffusion model parameters is highly innovative. Additionally, detailed explanations are provided for the introduced learnable low-rank parameters.",
"strength... | 7.5 |
pymXpl4qvi | [
"strength: The paper presents important findings, which open up future directions to resolve the issue.",
"strength: The section on adversarial attacks is praiseworthy, as it utilizes one of the findings to describe something very practical.",
"strength: Both findings are substantiated with some theoretical und... | 6 |
KD9F5Ap878 | [
"strength: I am impressed with the authors’ ambitious goal of covering a wide range of optimization applications. I have questions and concerns about how those applications and formulations were derived, by I applaud the attempt at building a comprehensive database.",
"strength: The algorithmic/theoretical contri... | 5 |
kT6oc5CpEi | [
"strength: The paper provides numerous visualizations, which serve as a valuable supplement to the textual content.",
"strength: The introduction of Multi-Objective concepts to enhance attack efficiency in this domain is innovative and represents a promising direction for future research.",
"strength: The use o... | 3 |
A0mk2Wi68Y | [
"strength: The goal of the paper, the presentation, and the implementation details are clear",
"strength: The approach does not require costly annotations for concepts and can (arguably) be applied to any time-series data",
"strength: The paper is well-written, and the idea is expressed clearly.",
"strength: ... | 5.5 |
SXTmAdGjlg | [
"strength: Originality: The authors proposed a adaptive optimization algorithm based on mirror descent for solving bilevel optimization problem without knowing the Lipschitz constants.",
"strength: Quality: Compared with approximation bilevel method, their approach is more robust and practical based on the numeri... | 4.6 |
4R71pdPBZp | [
"strength: This paper proposes a multi-agent collaboration approach to address software development problems.",
"strength: This paper introduces a software development benchmark called RSD-Bench, which provides more detailed and structured software requirements for documenting user needs compared to previous benc... | 7 |
uuriavczkL | [
"strength: As the paper points out, the counterfactual randomziation procedure in itself has been alluded to in previous work. But, the main contribution of this paper is a characterization of when a counterfactual distribution can be 'realized'.",
"strength: I liked that the paper operates under this structure f... | 7.5 |
p4RAKZ4oik | [
"strength: The paper addresses a novel problem space, combining federated learning (FL), black-box prompt tuning, and discrete prompt optimization.",
"strength: The authors target a more realistic setting in which the clients cannot access the embedding layer and the output distribution.",
"strength: The author... | 3 |
Q7EjHroO1w | [
"strength: The paper is written in a clear way and is well structured.",
"strength: The proposed method is formulated clearly and scoped within relevant literature.",
"strength: The proposed method is also evaluated experimentally, including comparisons with alternatives.",
"strength: Figures and visualisatio... | 6.5 |
37EXtKCOkn | [
"strength: The challenge seems well grounded as the sparse data over a large spatial domain is common for many types of problems, e.g., few agent trajectories over a large geographical domain.",
"strength: The method looks (possibly) scalable with low-resolution linear interpolation.",
"strength: The math formu... | 7.5 |
uqe5HkjbT9 | [
"strength: The multi-agent RL has the potential to solve more complex tasks. This paper aims to use multi-agent RL to solve the multi-goal task. The main contribution of this paper is by proposing a framework with doing clustering and prediction on the latent space and learning the policies based on the generated t... | 5.75 |
dsALpkd1OU | [
"strength: The paper tackles an important and interesting problem. Improving code agents is likely to have a large impact and be of significant interest to the community.",
"strength: The proposed approach is intuitively clear. The main ideas are easy to follow.",
"strength: There seems to be some algorithmic n... | 1.666667 |
qb2QRoE4W3 | [
"strength: The approach shows competitive performance on a variety of benchmarks, although performance is considerably lower on the more recent FreshQA dataset.",
"strength: The main strength of LLM-Cite is the cost savings compared to Google Search",
"strength: I have found the paper to be easy to read. The me... | 3 |
N18Z2MkMEa | [
"strength: A *combination* of RAG and MAML and application for code generation.",
"strength: The combination of 4 rewards beyond functional correctness to put into the RL pipeline.",
"strength: Empirically, comprehensive and solid experiments on various benchmarks.",
"strength: In detail, experiments (with Co... | 3 |
Nr6V30wK1l | [
"strength: I am drawn to the idea of extending the flow matching idea to other settings, such as conditional distributions. CVFM does so in a simple, straightforward way, which consists of smart sampling of source and target pairs, and conditioning of the neural network. The authors also extend the theory of FM and... | 4.5 |
EHYbqCDRtM | [
"strength: The authors seek to develop a fully interpretable method for graph representation learning, which a promising research direction.",
"strength: The authors provide open-source code for peer review.",
"strength: This paper introduces a new way to utilize LLMs for interpretable graph learning.",
"stre... | 2 |
huuKoVQnB0 | [
"strength: This paper focuses on avoiding costly pre-training ablations for LLM pre-training data selection and proposes a method that leverages public, pre-trained LLMs based on the correlation between downstream performance and perplexity. This aim is quite important for today's LLM practitioners and the proposed... | 6 |
KWo4w1UXs8 | [
"strength: Novelty. The model takes advantage of the inherent graph structure of human keypoints, allowing it to better capture spatial relationships and dependencies among keypoints. To my knowledge, GNN is not used in human-specific diffusion-based generation yet.",
"strength: The paper presents GUNet within th... | 3 |
ilOEOIqolQ | [
"strength: The paper presents a good empirical investigation on the creativity (and and lack of) of large language models compared to creative inputs from humans.",
"strength: The proposed Creative Index is simple and easy to understand.",
"strength: The paper is readable, contains quality visualizations to com... | 7 |
GsR3zRCRX5 | [
"strength: The paper is well-written and clearly structured.",
"strength: The figures are visually appealing, and the results are reported over a reasonable number of repetitions and with error bars.",
"strength: The introduction and background section are concise and easy to follow, except maybe for section 3.... | 6.166667 |
xFezgECSLa | [
"strength: The stated objectives of the paper are excellent if they can be achieved.",
"strength: The paper advocated analysis of LLM-based algorithms.",
"strength: The paper tackles an interesting and important problem.",
"strength: I like the proposed idea of analyzing the LLM-based algorithms as computatio... | 3 |
CPhqrV5Ehg | [
"strength: This paper shows a good analysis that RAD learns a low-rank approximation and therefore similar performance on toxicity/sentiments can be obtained with low-rank approximation. The reward model is split into two parts -- one for baseline reward estimation for the provided prefix and a separate term for th... | 6 |
bH6T0Jjw5y | [
"strength: The paper is nicely written and the method is explained well.",
"strength: The information theoretic approach is a good motivator for what follows in the paper. The method itself is going through a couple of departures from the pure mutual information and get increasingly complex and further approximat... | 8 |
JJH7m9v4tv | [
"strength: Using post-hoc discriminator guidance to enhance modeling performance is indeed a reasonable approach.",
"strength: This method effectively addresses sampling bias between the target density and the modeled density through the gradual correction enabled by annealed Langevin dynamics sampling. By iterat... | 3 |
P2Fjm0nIit | [
"strength: The paper is well-written and effectively communicates the proposed method and its technical details. The authors provide clear explanations of the neural compression based framework.",
"strength: The paper introduces a novel neural compression based framework for NeRF, which is a unique approach compa... | 4.333333 |
UHPnqSTBPO | [
"strength: This paper provides clear methodological contributions to improving human-LLM alignments.",
"strength: I especially appreciate the cascaded selective evaluation approach, which explicitly considers the costs of using stronger/more costly LLMs.",
"strength: Overall, this is an excellent paper with cle... | 8 |
KkrDUGIASk | [
"strength: Presents a highly intriguing new question, effectively addressing the challenges faced during the deployment of multi-agent collaborative perception systems.",
"strength: The approach in this paper is notably succinct and efficient; the authors design a novel backward alignment mechanism for individual... | 6.75 |
SZErAetdMu | [
"strength: The method is characterized by its simplicity and its clear and accessible presentation.",
"strength: The literature review is commendable, effectively contextualizing the research within its broader academic framework.",
"strength: The paper demonstrates a comprehensive approach to downstream tasks,... | 3 |
i1Yxnar4mj | [
"strength: The GGNN framework can systematically define and refine diverse covers of graphs from an algebraic perspective, which is novel and offers plenty of design space.",
"strength: Overall, the paper is well presented.",
"strength: I checked with the proofs and have not found fatal mistakes.",
"strength:... | 4.5 |
bXk9gcKhqp | [
"strength: This paper presents some theoretical and empirical results that will be of interest to the GNN community.",
"strength: The theoretical results are very simple: Theorem 1 is a standard result in algebra about polynomials, and Theorem 3.3 can be easily checked from first principles.",
"strength: The ma... | 4 |
D5X6nPGFUY | [
"strength: The paper introduces a novel Probabilistic Vision-Language Model (PrVLM), named ProLIP, which represents a shift from deterministic vision-language models to probabilistic embeddings. This approach innovatively captures the natural ambiguity in image-text relationships, where multiple captions may descri... | 6.4 |
sbG8qhMjkZ | [
"strength: This paper establishes quantitative convergence guarantees for finite-particle (and discrete-time) SVGD, with a much better dependency on problem parameters than the only previous known analysis. This is thus arguably the first satisfactory convergence bound, for an algorithm that has attracted considera... | 8 |
AmEgWDhmTr | [
"strength: The results are interesting and novel, tackling the important question of training dynamics of transformers and the role of chain-of-thought prompting.",
"strength: The motivation is good: understanding CoT is a significant direction to understand pre-training and finetuning LLMs, especially for reason... | 7 |
VSVQljJU5N | [
"strength: The paper extends Sheaf Neural Networks specifically to recommender systems, and experimental results show its effectiveness.",
"strength: I like the idea of hierarchical optimization of loss functions, as it prioritizes different constraints to enhance model robustness.",
"strength: The paper addres... | 3 |
88Qm4fGWzX | [
"strength: The proposed training-free approach is easy to adopt and can generate satisfying results.",
"strength: The paper is clear and easy to follow.",
"strength: The method seems correct",
"strength: The writing and organization seem clear",
"strength: The efforts to formalize a new benchmark and task, ... | 5 |
z4Ho599uOL | [
"strength: Novel Application of LLMs: This is the first work applying LLMs to JSSP, pushing the boundaries of LLM applications beyond traditional language processing tasks. The concept of fine-tuning an LLM on a scheduling problem is innovative.",
"strength: Dataset Contribution: The introduction of the Starjob d... | 3 |
yBIJRIYTqa | [
"strength: This paper proposes and studies a novel and interesting problem. It is pretty smart to consider the case where the cost of an arm could be negative.",
"strength: The results are rich and good. It studies multiple cases, and provides good results for them. The results look good enough from my understand... | 6 |
6E8GCcCgxl | [
"strength: The use of structured pruning in Eidetic Learning for addressing catastrophic forgetting is unique.",
"strength: This theoretically sound solution offers clear guarantees for preventing forgetting.",
"strength: Additionally, the authors plan to release a PyTorch framework, which will facilitate adopt... | 3.25 |
7BDUTI6aS7 | [
"strength: The paper provides a quite general connection between one generalized f-divergence DRO and the so-called fundamental risk quadrangle and applies to general cost functions.",
"strength: The FRQ framework provides interesting link between various problems in learning and risk management.",
"strength: T... | 3 |
7ha61H73pg | [
"strength: The proposed method can offer improvments for fine-tuning LLMs while saving memory usage.",
"strength: Extensive experiments are provided in this paper.",
"strength: The work describes a potentially useful method",
"strength: This paper is well-written and easy to understand. The proposed ILA metho... | 4.4 |
o8tjamaJ80 | [
"strength: This paper provides an interesting view of improving mixed sample qualities through adversarial training in the close-loop optimized mixup augmentation framework. The overall presentation of the manuscript is easy to follow, and the proposed methods are well-motivated.",
"strength: Extensive experiment... | 7.333333 |
ypBYdetYd9 | [
"strength: The study addresses an important question on diversity and degeneracy in RNNs.",
"strength: The authors investigate this question across multiple levels: behavior, dynamics, and connectivity, and across multiple settings: different loss functions, network sizes, and weight regularization methods.",
"... | 4.2 |
jxpsAj7ltE | [
"strength: Paper is well-written, easy to understand, and provided with source codes that model how Soft MoE is being implemented.",
"strength: Experiments are comprehensive and detailed with respect to its domain, with emphasis not only on performance, but also in terms of inference and wall-clock time.",
"str... | 7.5 |
YHDY5uXOSN | [
"strength: The studied problem is practically significant in communication systems",
"strength: The idea is quite original: the idea to use neural BP, while natural, requires adaptation of the original DDPM via VDM.",
"strength: The improvement seems substantial.",
"strength: The proposed method is well-motiv... | 3 |
ZGkfoufDaU | [
"strength: The originality of Min-K%++ lies in its focus on identifying local maxima to detect training data, moving beyond heuristic-based methods by leveraging theoretical insights from maximum likelihood training.",
"strength: The quality of the experiments is high, covering multiple benchmarks and various mod... | 7.5 |
3fl1SENSYO | [
"strength: The paper is well-written with a clear, thorough, and concise introduction that effectively summarizes key points from previous works",
"strength: The authors specifically address the challenges of the problem and provide clever solution to mitigate them",
"strength: The paper's main novelty is suppo... | 7.5 |
7fuddaTrSu | [
"strength: The authors shows a study of the RSME regarding surface air temperature and precipitation for single emulator among . PACE outperfoms the other emulators on 13 models on 15 for temperature and 9 models over 15 for precipitation. A box plot summarize the overall performances regarding these 15 models. PAC... | 3 |
uxFme785fq | [
"strength: The paper tackles an interesting though not entirely novel problem of releasing DP statistics from big-data that yield valid statistical inferences.",
"strength: The paper makes an effort to ensure the statistical properties of the algorithm are well understood which is important for applied researcher... | 2.5 |
hpCfPEvBsr | [
"strength: The evaluation provided in the paper is extensive, both considering the number of tasks and models evaluated. This paper can really become a reference for many future works.",
"strength: The tasks considered are handling very different modalities, which can foster the adoption of this dataset from diff... | 7.5 |
4VIgNuQ1pY | [
"strength: The authors provide a solid theoretical foundation for the proposed Neural SDEs, including the existence and uniqueness of solutions.",
"strength: The robustness section in the paper provides valuable insights into the proposed Neural SDEs' resilience against distribution shifts and input perturbations... | 6.666667 |
k2q0rUX2lx | [
"strength: The paper addresses a significant issue in the field of language model alignment and multi-step reasoning, which is a valuable contribution to the advancement of LLMs.",
"strength: The presentation of the material is coherent and the paper is generally easy to follow, which aids in the understanding of... | 3.5 |
IZiKBis0AA | [
"strength: FILTER attempted to add mechanistic insights by analysing of biochemical pathway data.",
"strength: The authors outline procedures for designing antibiotics using multiple machine learning models to predict various properties, identify related pathways through embeddings and clustering, and perform doc... | 3 |
Frhj9T7ihK | [
"strength: Paper is well-written and easy to follow",
"strength: The authors carry out a detailed qualitative study about how clinicians and experts could utilize such a model for diagnosis of a mental disorder",
"strength: The paper does address an important area of interest in psychiatry and AI.",
"strength... | 3 |
P7KIGdgW8S | [
"strength: Deep graph neural networks is an important area of research, and there are an enormous number of different architectures that have been proposed.",
"strength: This paper provides some valuable insights into which networks give better separation properties.",
"strength: The lower Holder estimates in e... | 8 |
tC1b9DBWww | [
"strength: I appreciate that the authors conducted extensive experiments to discuss bias and fairness in object detection.",
"strength: This fairness and bias topic is import for person detection. It could be beneficial to future model design and dataset construction in this research field.",
"strength: Extensi... | 2.5 |
rwdeKOdAwY | [
"strength: The usage of retrieval augmented strategy for classification.",
"strength: The paper claims to achieve state-of-the-art results across various settings; however, in my opinion, the comparison is unfair, as noted in my comments in the weakness section.",
"strength: [+] Pursuing better interpretability... | 3 |
70A6oo3Il2 | [
"strength: The proposed method is reasonable. Different samples has different difficulty. It is reasonable to use different loss.",
"strength: The paper is well-written and easy to follow.",
"strength: AdaFlood emerges as a trailblazing regularization strategy that sensitively tunes the flood level according to... | 4.333333 |
0bmGL4q7vJ | [
"strength: The idea of using an LLM to separately generate queries, files, and trajectories, followed by a query-file verifier and trajectory verifier is neat.",
"strength: The paper addresses the problem of using the appropriate array of tools corresponding to the information relevant to a query well.",
"stren... | 7.5 |
KBo7Z5aTV0 | [
"strength: The idea of investigating semantic segmentation from pixel feature distribution perspective is novel.",
"strength: The proposed pixel learning framework and its components are technically solid and innovative.",
"strength: The motivation and the underlying principles of designing the framework and ea... | 6 |
kxnoqaisCT | [
"strength: This paper contributes to reframes GUI interaction as a pure visual grounding problem, challenging the conventional wisdom that additional textual representations are necessary.",
"strength: The authors develop a novel way to generate diverse referring expressions (REs) by categorizing them into visual... | 7.75 |
lBwxmTbY6Z | [
"strength: The paper is well-written, it provides a unified yet concrete view of how to capture Granger Causality in a generative manner.",
"strength: Experimental results show consistent outperformance on both forecasting and anomaly detection.",
"strength: The paper posts a good review of literature for model... | 3.75 |
Dojny642Dy | [
"strength: The paper includes a questionnaire survey on user search behavior, highlighting users' preferences for interactive search functionalities. Moreover, it summarizes users’ intricate behavioral patterns that underscore the necessity of an interactive retrieval system.",
"strength: The proposal of the larg... | 4.666667 |
9vkgAaCI3F | [
"strength: The paper is clearly written for the most part, and fairly easy to follow.",
"strength: The RDAC framework is a neat idea to examine changes in the network weights / activations, and to analyse the stability-plasticity tradeoff during learning. The framework is simple enough that it can be applied to a... | 5.25 |
G0uhaIXmFw | [
"strength: It seems that the proposed Safe Low-Switching Primal-Dual Model-Based Algorithm achieves a regret bound that reduces the dependency on the state-action space and planning horizon, making it more effective in large-scale environments.",
"strength: A new approach with a low switch with improved bounds on... | 4.75 |
sDmjlpphdB | [
"strength: The motivation is interesting and reasonable. Typically for a downstream task, a fixed task instruction sentence may not be optimal to LLM for every sample of this task. In most cases, a downstream task dataset can also be subdivided into subtasks requiring various instructions. From these perspective, a... | 4.75 |
KjTh5C0z7Y | [
"strength: The work considers a novel framework of converting a truth table into an AIG instead of optimizing the synthesis process.",
"strength: Combining the supervised learning of a transformer-based network and the RL is also new, and the results look promising.",
"strength: The paper is well written and ea... | 4.5 |
Hm7RYDspQP | [
"strength: The paper is well motivated and proposes an interesting approach to tackle an important issue in robotics.",
"strength: Considering alternative algorithms to DRL which are more sample-efficient for specific applications is a timely and important problem.",
"strength: The evaluation does support the c... | 3.5 |
aJ3tiX1Tu4 | [
"strength: The paper rigorously explores conformal prediction under complex distribution shifts, going beyond covariate shift to address both covariate and concept shifts.",
"strength: The decomposition of the Wasserstein distance to measure and address different types of shifts is novel and provides a clearer st... | 6.666667 |
c5pwL0Soay | [
"strength: The empirical study of this paper is very sound and solid. The paper evaluates the method on various control tasks, including locomotion and manipulation tasks. Besides, the paper aims to address the unsupervised RL problem on visual-based tasks, which are much more challenging in the area. The paper als... | 7.5 |
UFwefiypla | [
"strength: The only strength of this paper is the idea of leveraging a text-based language model to improve audio codec, which is reasonable and interesting.",
"strength: The proposed DM-Codec method improved the existing speech tokenization system which only incorporates acoustic and semantic information (e.g Sp... | 3 |
izDiFGXn9B | [
"strength: If it holds up, the core result— that comparing just the weights of models often performs better on statistical testing for functional similarity than representation similarity— is both interesting and surprising. The method also is substantially more efficient, does not require data, and may be compleme... | 3.5 |
s20W12XTF8 | [
"strength: How to defend against jailbreak attacks is still an open challenge in the community.",
"strength: Multiple LLMs are evaluated.",
"strength: The proposed attack does not rely on the attacks. Moreover, the method is efficient as it directly manipulates the internal states of an LLM during its inference... | 6.25 |
pqOjj90Vwp | [
"strength: Solid Theory: The paper constructs a solid theoretical foundation through extensive mathematical proofs and derivations.",
"strength: Unified Framework: By establishing a connection between GNNs and first-order logic, the paper proposes a theoretical framework that can unify different GNN models. This ... | 8 |
4lOWCkhr4g | [
"strength: The design of the experiment is comprehensive. The source & target language set covers not just common European languages but also the less common Arican languages such that the cross-family source & target language situation can also be studied.",
"strength: Comparison with baseline systems such as su... | 5.25 |
fxQiecl9HB | [
"strength: The introduction effectively describes the gap in the literature it addresses, namely the periodic nature of crystal structures, which is not considered in the core of the GNN literature devoted to molecular modelling. The paper provides a comprehensive review of graph and transformer models for related ... | 7.25 |
XrunSYwoLr | [
"strength: This paper overcomes the differences in computational paradigms between ANN and SNN Transformers, and can accurately approximate ANNs with converted models.",
"strength: The proposed training-free conversion strategy could enable the direct deployment of large-scale pretrained ANN models to low-power n... | 7 |
DFTHW0MyiW | [
"strength: This paper proposes an interesting approach for robust reinforcement learning by adaptively switch among non-dominated policies.",
"strength: The definition of \"non-dominated policy\" seems new.",
"strength: The algorithm to discover $\\tilde{\\Pi}$ is non-trivial and might be considered in future s... | 7 |
4GfEOQlBoc | [
"strength: the work is well-grounded in the theoretical vision science field with sufficient references to previous research",
"strength: extensive model comparison (several predictive models vs several perceptual distances)",
"strength: validation on empirical data",
"strength: An interesting study on \"Dise... | 5.25 |
wHLDHRkmEu | [
"strength: The proposed method brings a few trainable parameters into CLIP for both feature adaptation and modal fusion, which achieves the best performance on several RIS datasets.",
"strength: The proposed method utilizes a novel intertwined structure to assist modal feature fusion.",
"strength: The paper is ... | 7.333333 |
dVq2StlcnY | [
"strength: The paper offered some interesting insights into interpretable GNNs and proposed some solutions. The experiments on the proposed methods are extensive and shown pretty good results.",
"strength: It's interesting to see some works focusing on better understanding the theoretical foundation for the expre... | 5.75 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.