Tracked topic
…Jiahao Chen , , , , , , Abstract EDMolGPT is a decoder-only autoregressive framework that generates molecules from low-resolution electron density point clouds, leveraging physically meaningful density signals to produce structurally accurate 3D conformations. AI…
…when all sampled traces of a given prompt receive identical rewards, group-relative advantage estimation provides no gradient signal, even though the traces may differ substantially in reasoning quality. We propose Reasoning…
…We further show that Z-Reward can serve as a differentiable reward signal for text-to-image optimization , yielding a 41.3% net human-preference improvement over the SFT baseline. View arXiv…
…AI-generated summary Agentic reinforcement learning (RL) for Large Language Models (LLMs) critically depends on the exploration capability of the base policy , as training signals emerge only within its in-capability region…
…it likely helps with clean on-policy signals, but if experts disagree on an example, the student might miss useful guidance and hurt robustness, especially in composition-heavy or out-of-domain…
…These results identify geometry conflict as both an explanatory signal for forgetting and a practical control signal for LLM continual post-training . View arXiv page View PDF GitHub 5 Add to collection…
…The resulting framework further supports iterative self-evolution , allowing the model to improve from unlabeled remote sensing imagery through its own verification signal. Experiments show that RemoteZero achieves competitive performance against strong…
…a stale artifact is not just delayed work, but readable evidence that the LLM can reflect on, revise, and turn into useful evolution signal. FlashEvolve further improves throughput and token efficiency with…
…We introduce BraveGuard, a self-evolving defense framework for training guard models from open-world threat signals and realistic agent trajectories . BraveGuard mines recent research sources to identify emerging risks and attack…
…AI-generated summary Evaluation of software engineering (SWE) agents is dominated by a binary signal: whether the final patch passes the tests. This outcome-only view treats a principled solution and a…