Self-Policy Distillation extracts a capability subspace from model gradients on correctness tokens, projects KV activations into it for self-generation, and fine-tunes LLMs to achieve up to 13-16% gains over baselines without external signals.
hub
arXiv preprint arXiv:2312.06585
17 Pith papers cite this work. Polarity classification is still indexing.
17
Pith papers citing it
hub tools
citation-role summary
background 3
citation-polarity summary
roles
background 3polarities
background 3representative citing papers
POPO uses bounded importance sampling on positive rollouts and a siamese policy network to achieve implicit negative gradients and stable optimization, matching or exceeding GRPO on math benchmarks such as 36.67% on AIME 2025.
SAPO introduces segment-level policy optimization using a step-wise MDP abstraction to better align RL updates with reasoning structure in multi-modal LLM tasks.
SSPT turns space-syntax integration metrics into post-training feedback signals that improve public-space dominance and functional hierarchy in AI-generated residential floor plans.
rePIRL learns effective process reward models for LLM reasoning via a dual policy-PRM update process inspired by inverse RL, unifying online and offline methods with reported gains over prior approaches on math and coding datasets.
Iterative SFT-RL cycles enable a 7B LVLM to develop sophisticated visual chain-of-thought reasoning and improve performance on math and general reasoning benchmarks.
Process advantage verifiers trained to predict step-level progress under a distinct prover policy improve LLM reasoning accuracy by over 8% and sample efficiency by 5-6x over outcome reward models.
SCoRe uses multi-turn online RL with regularization on self-generated traces to improve LLM self-correction, achieving 15.6% and 9.1% gains on MATH and HumanEval for Gemini models.
SOLAR introduces a self-optimizing agent using meta-learning on model weights and RL-driven strategy discovery for lifelong adaptation in LLMs, claiming superior performance on reasoning tasks across domains.
ITEM is a new iterative utility judgment loop for RAG that maps Schutz's three levels of relevance to retrieval, utility scoring, and generation, yielding measured gains on TREC DL, WebAP, GTI-NQ, and NQ.
PRISMA augments self-training with direct preference optimization and an emotion-aware negotiation strategy chain-of-thought to produce more interpretable and effective negotiation dialogues on two new datasets.
Experiments show that long CoT reasoning in LLMs emerges with more training compute when reward shaping is used properly, and scaling verifiable rewards from noisy data helps especially on out-of-distribution tasks.
The survey organizes the shift of LLMs toward deliberate System 2 reasoning, covering model construction techniques, performance on math and coding benchmarks, and future research directions.
A survey that organizes LLMs-as-judges research into functionality, methodology, applications, meta-evaluation, and limitations.
This survey frames foundation agents using brain-inspired modular architectures and reviews challenges in evolution, collaboration, and safety.