FOCUS: PhD project tackles credit assignment for long-horizon agents with active critic learning
EliasEskin · x · 2026-11-03
Zhengping Jiang's PhD final project FOCUS tackles a core challenge in training long-horizon agents: they can take hundreds of actions before receiving a single outcome reward, raising two questions — which actions deserve credit, and how training fits with existing compaction and memory strategies.
FOCUS performs active critic learning over managed context states, with modular, customizable context protocols and integrations with Tinker and slime. Documentation is public.
The author is on the job market; multiple academics (including his advisor of 7 years) endorsed the work as highly relevant to many labs and businesses.
More from coding & agent
- Open-Source MCP Server Demo Hooks Custom Tools into OpenAI via Cloudflare Tunnels — modelcontextprotocol · 2026-11-03
- Jam Launches MCP Server Giving AI Agents Instant Bug Context from Report to PR — modelcontextprotocol · 2026-11-03
- Task boundaries matter: dev gets major quality gains breaking Claude work into small tasks — LadyAshBorg · 2026-11-03
- NVIDIA's AgentBaton cuts agent API costs 65.9% via well-timed handoffs to smaller models — rupspace · 2026-11-03
- Deepgram acquires Daily; open-source voice agent framework Pipecat to stay open with a foundation — sven_ai · 2026-11-03
- AI feature usage is up but profit is down: audit these five cost sinks first — djgoel123 · 2026-11-03