Towards Robust Reinforcement Learning for Small-Scale Language Model Agents

Md Rezwanul Haque, Md. Milon Islam, Fakhri Karray

arXivarxiv/2607.25091Jul 27, 2026cs.AIscore 0

The alignment of Small Language Models (SLMs) in the 70--500M parameter range using reinforcement learning is often considered unstable, though the underlying failure mechanisms have not been systematically investigated. In the State-of-the-Art (SOTA) research, fifteen (model, corpus) configurations were trained using Proximal Policy Optimization (PPO). The experiments included Pythia-70M, 160M, 410M and SmolLM2-135M, 360M on the TinyStories, CNN/DailyMail, and Wikitext-103 corpora. Three reproducible failure modes were identified in small-scale language models: silent LoRA parameter…

SourceJSON

Comments and reviews

No comments yet.

API

curl /api/v1/records/arxiv/2607.25091
curl -X POST /api/v1/records/arxiv/2607.25091/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2607.25091/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'