Training Small Language Models for Transferable Tool Selection (arXiv)
rohanpaul_ai · x · 2026-10-24
arXiv page for the Meta/CMU paper (arXiv:2610.31483) by Zhaoyi Zhou et al., submitted Oct 22, 2026: a Qwen3-8B post-trained with their RL recipe outperforms 402B Llama-4-Maverick on ToolBench tool selection and transfers directly to different retrievers and new-domain tool corpora.
Related event: Meta and CMU Train 8B Model to Beat 402B Llama in Tool Selection(3 posts)→
More from Research
- Reviewer chimes in: 'I once reviewed a paper with a negative R²' — RexDouglass · 2026-10-24
- Ropedia A0: code-only robot policy hits 70.24% on RoboDojo, 75% first-try sim2real — liuziwei7 · 2026-10-24
- Holi-World turns egocentric video into 3D/4D data: 4B model beats zero-shot 27B — liuziwei7 · 2026-10-24
- DecisionFormer open-sourced: fine-tune calibrated decision models on any LLM backbone — kalyan_kpl · 2026-10-24
- RLCR trains LLM agents as organizations: emergent hierarchies and division of labour — algo_diver · 2026-10-24
- Simplified Matching Distillation: few-step diffusion without adversarial training, validated up to 33B — ArashVahdat · 2026-10-24