Claims Browser

Recent post-training methods like RL and OPSD have driven rapid progress in mathematical reasoning for large language models

"Recent prominent post-training methods, such as Reinforcement Learning (RL) and On-Policy Self-Distillation (OPSD), have driven rapid progress in mathematical reasoning for large language models"

Process quality, result quality, and data representativeness are effective criteria for selecting high-quality agent training trajectories

"the first stage performs trajectory-level screening based on process quality, result quality, and data representativeness, selecting a high-quality and representative subset of successful trajectories."

Multi-granularity data selection that filters at trajectory and segment levels improves training efficiency and model performance for software engineering tasks

"training on the 10% trajectory subset selected by SWE-Prime outperforms training on the full resolved dataset, yielding relative performance gains of up to 12.2% and 24.2%, respectively."

Training on successful agent trajectories can introduce noisy supervision because successful trajectories may contain ineffective, redundant, or risky steps

"task success does not guarantee high-quality supervision: successful trajectories may still contain ineffective, redundant, or risky steps. Directly using such trajectories for SFT can introduce noisy supervision and encourage models to imitate undesirable problem-solving behaviors."

LLMs in code review exhibit critical weaknesses such as cross-round temporal misalignment and inadequate long-range memory

"our in-depth error analysis dissects the distinct drivers of false positives and false negatives, revealing critical weaknesses such as cross-round temporal misalignment and inadequate long-range memory"

LLMs' performance in code review varies substantially across different defect types and severity levels, with semantically complex or low-salience defects being significantly more likely to be missed

"LLMs' performance varies substantially across different defect types and severity levels, with semantically complex or low-salience defects being significantly more likely to be missed"

Mainstream LLMs exhibit limited overall performance in defect detection and defect lifecycle state tracking, with performance degrading significantly as the number of interaction rounds increases

"experiments reveal that mainstream LLMs exhibit limited overall performance in defect detection and defect lifecycle state tracking, with performance degrading significantly as the number of interaction rounds increases"

Most existing LLM approaches to automated code review oversimplify code review into a single-round, static decision task, failing to capture the multi-round interactive nature of realistic review scenarios

"Although recent work explores large language models (LLMs) for automated code review, most approaches oversimplify code review into a single-round, static decision task, which fails to capture the multi-round interactive nature and the complex problem-solving processes inherent in realistic review scenarios"

RedEvoAgent outperforms fixed and agentic baselines on multiple benchmarks, target models, and execution harnesses while improving tool efficiency and transferring across attacker models and target execution harnesses

"Experiments on multiple benchmarks, target models, and target execution harnesses show that RedEvoAgent outperforms fixed and agentic baselines, improves tool efficiency, and transfers across attacker models and target execution harnesses."

Trajectory-based retrieval in agentic attackers can reuse misleading experiences due to retrieval bias and unclear tool credit, with full trajectories adding context overhead while reducing interpretability

"such retrieval can reuse misleading experiences due to retrieval bias and unclear tool credit, and full trajectories add context overhead while reducing interpretability"

Recent agentic attackers that coordinate multiple jailbreak tools and use trajectory-based retrieval show stronger potential than fixed attack methods

"recent agentic attackers coordinate multiple jailbreak tools and show stronger potential through trajectory-based retrieval"

LLM-based agents deployed in product-level execution harnesses create greater risks through jailbreaks triggering harmful tool use and persistent state changes than unsafe text generation alone

"LLM-based agents are increasingly deployed in product-level execution harnesses, where jailbreaks can trigger harmful tool use and persistent state changes, creating greater risks than unsafe text generation alone."

Most machine learning approaches model chemical reactions through de novo generation or heuristic graph edits rather than through electron space transformations

"most machine learning approaches model them either through \textit{de novo} generation of product molecules or through heuristic graph edits that operate directly on molecular topology."

MAELLE naturally recovers mechanistic trajectories that align with known chemistry and can predict side products of a reaction

"because the learned flow operates over the full electron redistribution, MAELLE naturally recovers mechanistic trajectories that align with known chemistry and can predict side products of a reaction."

MAELLE maintains strong performance in out-of-distribution settings where existing methods degrade

"we evaluate robustness across two out-of-distribution settings - structural complexity and reaction type - and find that MAELLE maintains strong performance where existing methods degrade."

MAELLE achieves competitive performance on the USPTO-480K benchmark compared with leading reaction prediction models

"MAELLE achieves competitive performance on the USPTO-480K benchmark compared with leading reaction prediction models."

On DNA-to-amino-acid transduction, the method reduces runtime by several orders of magnitude compared to threshold-pruned beam summing and makes estimating prefix probabilities for long target strings feasible

"On a DNA-to-amino-acid transduction, it reduces runtime by several orders of magnitude relative to threshold-pruned beam summing and makes estimating prefix probabilities for long target strings feasible."

The proposed beam-summing algorithm achieves better compute-variance tradeoff on text and lower error on DNA compared to sequential Monte Carlo baselines

"We evaluate the method on encyclopedic text and DNA against sequential Monte Carlo baselines that resample with replacement. It achieves a better compute--variance tradeoff on text and lower error at the same maximum number of particles on DNA."