vivo AI Lab's CAPO-SLT fixes RL drift in sign language translation with confidence-aware clipping
量子位 · wechat · 2026-09-09
vivo AI Lab introduces CAPO-SLT, addressing why sign language translation models that align vision and text well still drift during autoregressive generation. In standard PPO/GRPO, all positive-advantage tokens share one clipping limit, over-reinforcing high-confidence language-prior "fluent but wrong" tokens while suppressing low-confidence but genuinely visual-grounded ones. CAPO sets the upper clip bound per token based on the old policy's confidence (0.2 base, relaxed up to 0.3), plus a loss cap for negative-advantage tokens — without changing the visual backbone or reward.
Trained with supervised initialization on ST-GCN pose features (face/body/two hands) then RL with sentence-level rewards combining BLEU-1, BLEU-4, ROUGE-L and BERTScore, CAPO-SLT uses only pose input yet beats Geo-Sign (+2.93/+1.26/+0.64) and Uni-Sign (+4.96/+3.07/+3.67) on CSL-Daily, topping even Pose+RGB methods, and reaches 41.4 BLEU-1 on How2Sign. Ablations show +0.30/+0.50/+0.67 over DAPO under identical reward and settings, with more stable policy entropy during training.
Failure examples without CAPO include translating "I introduce these colleagues" as "I'm a team of several people" — fixed after adding CAPO. Limitations: milder ASL gains and reference-based rewards that under-cover valid paraphrase. The approach is low-intrusion for existing PPO/GRPO pipelines. Paper on OpenReview.
More from Research
- 2-billion-line Lean proofs: mathematicians debate proof without understanding — Singularitarian · 2026-09-09
- OpenAI: 10,000 AI agents produced a Navier–Stokes solution in 88 hours, sparking credit dispute — Wes Roth · 2026-09-09
- A digital fruit fly brain can play Beat Saber — cixliv · 2026-09-09
- KVMem pages agent context overflow to KV state, beating compaction on DeepSWE — omarsar0 · 2026-09-09
- Collision attacks on SHA-2 pushed to 39 steps in new paper — jedisct1 · 2026-09-09
- 20,000-qubit trapped-ion quantum computer could break 256-bit ECC in 26 days — jedisct1 · 2026-09-09