11 papers
GradCuit: Credit-Assigned Gradient Flow Enables Robust and Interpretable Test-Time Latent Reasoning
Zhaoxin Yu, Qi Shen, Hengli Li +4
Optimization-based latent reasoning improves large language model outputs by optimizing instance-specific continuous states at test time while keeping model parameters frozen. Exis…
The Price Reversal Phenomenon: When Cheaper Reasoning Models Cost More
Lingjiao Chen, Chi Zhang, Yeye He +3
Developers and consumers increasingly choose reasoning models (RMs) based on their listed API prices. However, how accurately do these prices reflect actual inference costs? We con…
Metis: Learning to Jailbreak LLMs via Self-Evolving Metacognitive Policy Optimization
Huilin Zhou, Jian Zhao, Yilu Zhong +7
Red teaming is critical for uncovering vulnerabilities in Large Language Models (LLMs). While automated methods have improved scalability, existing approaches often rely on static…
Seirênes: Adversarial Self-Play with Evolving Distractions for LLM Reasoning
Chi Zhang, Haibo Qiu, Qiming Zhang +3
We present Seirênes, a self-play RL framework that transforms contextual interference from a failure mode of LLM reasoning into an internal training signal for co-evolving more re…
Preserve Support, Not Correspondence: Dynamic Routing for Offline Reinforcement Learning
Zhancun Mu, Guangyu Zhao, Yiwu Zhong +1
One-step offline RL actors are attractive because they avoid backpropagating through long iterative samplers and keep inference cheap, but they still have to improve under a critic…
FAPO: Flawed-Aware Policy Optimization for Efficient and Reliable Reasoning
Yuyang Ding, Chi Zhang, Juntao Li +2
Reinforcement learning with verifiable rewards (RLVR) has emerged as a promising paradigm for enhancing the reasoning capabilities of large language models (LLMs). In this context,…