VETTA uses a shared Transformer critic with separate turn-value and token-value heads.
Coordinating turn- and token-level credit boosts multi-turn LLM agent training
VETTA learns both credit types via separate value heads on a lightweight shared critic, improving success rates on two agent benchmarks.
Top University
Jiaju Chen · Min Yang · Jinghua Piao · Xiaochong Lan · Xu Xia · Xiangnan He · +1 more
University of Science and Technology of China · Zhongguancun Academy · Shandong University · Tsinghua University · Southeast University
Research Digest··2 min read
The authors introduce VETTA, a credit assignment method that jointly models turn- and token-level advantages for multi-turn LLM agents.
Why this paper
From Tsinghua University and 4 others
In one line
VETTA jointly learns turn- and token-level credit via separate heads on a shared lightweight critic, improving multi-turn LLM agent success rates.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ✓Reports numbers on named benchmarks (2 benchmarks)
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§