IF-MAP uses influence functions to rank which agents to improve first in multi-agent LLM systems
Jie Mao · hf · 2026-10-24
Researchers from CUHK-Shenzhen, HIT and Harvard introduce IF-MAP, a method for deciding which agent to improve first in multi-agent LLM systems when upgrading every agent is prohibitively costly. Failure attribution alone doesn't reveal system-level gains from improvement.
For each agent, IF-MAP measures end-to-end loss changes under controlled semantic prompt perturbations, fits a regularized local response model in a low-dimensional prompt-embedding subspace, and uses the estimated response norm as a priority score. The black-box procedure requires no access to model parameters or gradients. Code: github.com/s1412/IF-MAS.
More from coding & agent
- Dev experiments with autoencoders over functions; low-dim search may be key to local FunSearch — cephaloform · 2026-10-24
- Agent with SSH access successfully upgrades an ancient PHP 5.6 server end-to-end — DanielLockyer · 2026-10-24
- DeepMind researcher demos a playable 3D racing game built right on her desk — thursdai_pod · 2026-10-24
- Cut LLM costs: route requests, disable thinking, cache answers, escalate only on low confidence — colinmcnamara · 2026-10-24
- NVIDIA's OpenShell: Apache-2.0 runtime sandbox for autonomous agent fleets, 18.6k stars — petrusenko_max · 2026-10-24
- Claude Code Routines runs overnight, screens 380 clinics into 30 outreach drafts — Arindam_1729 · 2026-10-24