Agent action prediction on otoSpeech 113-session (test)
70.9wF1DualTurn (E – Full FT)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DualTurn (E – Full FT)variant=E, finetuning=Full2026.03 | 70.9 | 49.8 | 86 | |
| DualTurn (A – LoRA (main))variant=A, finetuning=LoRA2026.03 | 70.7 | 51.2 | 84.8 | |
| VAP (LR-6)aggregator=logistic regression2026.03 | 46.1 | 0 | 71.9 |