Claims Browser

The proof technique separates two stability mechanisms: a global comparison argument based on order monotonicity and W_∞ contraction brings an arbitrarily initialized iterate into a local neighborhood, where linearization and martingale analysis apply.

"The proof separates two stability mechanisms. A global comparison argument, based on the order monotonicity of reward cumulative distribution functions and the $W_\infty$ contraction of the distributional Bellman operator, brings an arbitrarily initialized iterate into a local neighborhood."

The result sharply distinguishes between local stochastic fluctuation and global sample complexity in distributional RL, with the deterministic transient and burn-in depending on the smallest Bellman-target density (order m^(-1) in worst case).

"The result therefore distinguishes sharply between the local stochastic fluctuation and the global sample complexity."

ES can lead to broader reasoning coverage than GRPO, thereby better exploiting the reasoning capabilities of pretrained LLMs

"this paper first identifies a performance advantage of ES over GRPO, theoretically and empirically showing that ES can lead to broader reasoning coverage, thereby better exploiting the reasoning capabilities of pretrained LLMs."

Verifier-projected Jensen-Shannon diversity across the ES population is helpful to higher Pass@K performances

"Theoretically, we show that verifier-projected Jensen-Shannon diversity across the ES population is helpful to higher Pass@K performances."

ES improves Pass@1 while attaining higher Pass@K than GRPO, unlike GRPO which exhibits entropy collapse

"Empirically, unlike GRPO, which exhibits entropy collapse, ES improves Pass@1 while attaining higher Pass@K than GRPO."

The task-performance gains of ES are only contributed to a sparse subset of larger-magnitude updates despite substantial whole-model parameter drift

"we find that despite substantial whole-model parameter drift, the task-performance gains of ES are only contributed to a sparse subset of larger-magnitude updates."

Large parameter movement in ES need not imply widespread functional change and does not necessarily lead to catastrophic forgetting

"This functional sparsity suggests that large parameter movement need not imply widespread functional change, and held-out evaluations further show that it does not necessarily lead to catastrophic forgetting."

ES requires a smaller population size in a larger LLM

"we study how hyperparameter design affects the effectiveness of ES, demonstrating that ES requires a smaller population size in a larger LLM."

ES should be positioned as a distinct reasoning post-training paradigm rather than a less effective, memory-efficient alternative to GRPO

"These findings position ES as a distinct reasoning post-training paradigm rather than a less effective, memory-efficient alternative to GRPO."

Chain-of-thought reasoning can be used to improve the reliability of LLM responses

"Large Language Models (LLMs) can be trained to perform chain-of-thoughts reasoning in order to improve the reliability of their responses."

Fragment-based reasoning using parallel source-target fragments from similar exemplars can improve LLM-based machine translation

"We introduce a novel fragment-based reasoning framework in which the model first extracts parallel source-target fragments from retrieved similar exemplars, and uses these fragments as intermediate reasoning traces to produce the final translation."

Fragment-based machine translation significantly outperforms standard k-shot and basic drafting methods across 6 languages and up to 5 domains per language

"Our experiments with the Qwen3 model family, over 6 languages, including up to 5 domains per language, demonstrate that fragment-based MT significantly outperforms alternative methods like standard k-shot or basic drafting."

Batch prompting makes LLM inference more efficient by processing multiple instances simultaneously but suffers from unpredictable downstream task performance

"batch prompting makes large language model inference more efficient by processing multiple instances simultaneously, it suffers from unpredictable downstream task performance"

Cascaded batch prompting resolves the unpredictability of conventional batch prompting by disentangling complex reasoning from symbol grounding

"cascaded batch prompting, a two-stage approach designed to resolve the unpredictability of conventional batch prompting by disentangling complex reasoning from symbol grounding"

Cascaded batch prompting outperforms standard single prompting baseline while achieving speedup proportional to batch size, establishing new state of the art on the Pareto frontier

"the proposed method outperforms the standard single prompting baseline while achieving a speedup proportional to batch size, establishing a new state of the art on the Pareto frontier"

RL-style post-training can substantially improve chain-of-thought reasoning, long-horizon planning, and self-correction in language models

"such RL-style post-training ("RL-for-LLMs") can substantially improve chain-of-thought reasoning, long-horizon planning, and self-correction"

State-of-the-art reasoning language model training requires millions of GPU-hours

"state-of-the-art RLM training requires millions of GPU-hours"

RLM training is as much a parallel and distributed systems problem as an algorithmic one

"This makes RLM training as much a parallel and distributed systems problem as an algorithmic one"