The authors begin with an initial model-generated trajectory and optimize a short prefix of its hidden states without changing the model's parameters.
Token-aware latent refinement improves vision-language reasoning at inference time
The method directs visual and reasoning feedback to different hidden-state tokens while keeping model parameters frozen.
Chinese Tech
Hao-Xuan Ma · Yihao Liu · Yutao Sun · Yanting Miao · Mengyu Zhou · YiCheng Xiao · +5 more
Qwen Business Unit of Alibaba · Nanjing University · National Key Laboratory for Novel Software Technology, Nanjing University · Zhejiang University · University of Waterloo
Research Digest··2 min read
Ma et al.
Why this paper
From Chinese Academy of Sciences and 7 others · Released code
In one line
Token-disentangled latent test-time scaling improves vision-language reasoning by routing visual and reasoning feedback to different token roles.
What it released
Code
What we could check
- ✓Code link in the paper (github.com)
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ✓Reports numbers on named benchmarks
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§