LatentGym is introduced: a controllable suite in which each environment is organized around a ground-truth latent variable governing the structure across tasks, and a controlled foundation for studying how LLM agents learn from experience across tasks, and for designing agents that adapt more reliably in sequential, pe...
Daksh Mittal, Tommaso Castellani, Thomson Yen et al.· arXiv.org· 2 citations
This work introduces Seq@K, a natural evaluation protocol that measures whether a model can solve a problem within K sequential attempts, provided feedback after each previous failure, and establishes how Seq@K offers a more faithful metric for intelligent systems that typically operate in such feedback loop, and the n...
Thomson Yen, Gilbert Yang, Nai-Meng Ye et al.· 0 citations
This work introduces LOTS (Likelihood-Only Tool Scoring), which evolves an agent's tool space from accumulated output experience while keeping model parameters fixed and demonstrates that task-specific spaces persist and continue to improve over time, while cross-model experiments show that learned configurations trans...
Xuanqi Zhang, Rui-Nan Jin, Run Yang et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
