We have 11 of 185 papers
We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.
Not the right person? Other researchers publish under this name.
This work introduces a multi-agent framework designed to perform problem evolution while validating the solvability and increased difficulty of the generated problems, and provides empirical evidence that code-driven agents can serve as a viable mechanism for synthesizing high-difficulty mathematical reasoning problems...
This work identifies and formalizes this failure mode, which is term destructive resource preemption: obtaining the resources required for a requested task by terminating, overwriting, evicting, or degrading an incumbent task.
Temperature-Scaled On-Policy Self-Distillation (TS-OPSD), a lightweight policy reheating method that internalizes the exploratory effect of temperature into model parameters, suggests that entropy restoration can serve as a simple post-collapse intervention for extending reasoning-oriented RL.
This work proposes ISER (Isolation-based Spherical Ensemble Representations) that extends existing isolation-based methods by using hypersphere radii as a monotonic transformation of local density characteristics while maintaining linear time and constant space complexity w.r.t. the dataset size.
It is found that effective warm-up relies on teacher-compatible chain-of-thought supervision, and that even incorrect teacher rollouts can provide comparable benefits to correct ones, which suggests that warm-up primarily transfers a teacher-compatible thinking pattern rather than merely correct answers.
Terminal-Universe is introduced, a framework which turns each trajectory into a reusable environment and explores it for synthesizing new tasks and continued interactions, and produces 37.3k task-sufficient environments.
This work proposes to use bandwidth-calibrated MIP coupled with Tukey IQR peak-detection to isolate reasoning-crucial tokens at the output layer, and applies a Jaccard stability metric over multi-domain problems to verify if the MIP-identified tokens are reasoning quality-guaranteed.
This paper examines whether synthetic training artifacts can support the complete SFT-to-RL cycle for competitive programming, and synthesizes tasks, verified solutions, and reliable test cases that serve as reward signals for reinforcement learning.
This work introduces credit-addressable reasoning, in which the semantic units exposed during inference also define where learning compares alternatives and assigns credit, and instantiates Code-CoT, which retains the diagram, represents visual relations as line-addressable executable code, and organizes reasoning into...
Taking semantic segmentation task as an example, extensive experiments demonstrate that AirMoE outperforms MoE baselines and single-model competitors.
This work proposes AgentOPSD, a critic-free, recursive method for turn-level credit assignment in agentic reinforcement learning that aggregates token-level teacher-student log-probability gaps into turn-level evidence and recursively updates a Bayesian belief state in log-odds space.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
