CHASE uses co-evolutionary RL with GRPO to harden LLMs against black-box prompt-rewriting attacks, cutting mean StrongREJECT scores by 43.2% on held-out families while keeping zero false refusals on benign prompts.
arXiv preprint arXiv:2405.15624 , year =
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4representative citing papers
Pre-training provides a geometric warm start in a single-index model that enables weak-to-strong generalization up to a supervisor-limited bound, with empirical phase-transition evidence in LLMs.
SPS interleaves RL and IRL to counteract probability squeezing in LLM reasoning trajectories, improving Pass@k on five benchmarks while identifying an empirical upper bound on multi-sample performance.
rePIRL learns token-level process rewards for LLM reasoning via a guided-cost-learning-style IRL objective, and shows these rewards improve reasoning policies on math/coding benchmarks.
citing papers explorer
-
CHASE: Adversarial Red-Blue Teaming for Improving LLM Safety using Reinforcement Learning
CHASE uses co-evolutionary RL with GRPO to harden LLMs against black-box prompt-rewriting attacks, cutting mean StrongREJECT scores by 43.2% on held-out families while keeping zero false refusals on benign prompts.
-
On the Blessing of Pre-training in Weak-to-Strong Generalization
Pre-training provides a geometric warm start in a single-index model that enables weak-to-strong generalization up to a supervisor-limited bound, with empirical phase-transition evidence in LLMs.
-
SPS: Steering Probability Squeezing for Better Exploration in Reinforcement Learning for Large Language Models
SPS interleaves RL and IRL to counteract probability squeezing in LLM reasoning trajectories, improving Pass@k on five benchmarks while identifying an empirical upper bound on multi-sample performance.
-
rePIRL: Learn PRM with Inverse RL for LLM Reasoning
rePIRL learns token-level process rewards for LLM reasoning via a guided-cost-learning-style IRL objective, and shows these rewards improve reasoning policies on math/coding benchmarks.