NBPO on TL;DR at the selected eta_t
The TL;DR NBPO arm retrained at the eta_t the pooled-entropy rule selects (0.05) instead of the released eta_t = 1, reusing the same precomputed pair artifact; only eta_t differs.
- Panel: TL;DR
- Objectives: coverage, faithfulness, conciseness, helpfulness
- Backbone / reference policy:
meta-llama/Llama-3.1-8B-Instruct - Training budget: 300 optimiser updates, global batch 16
- Reported in: Nash Bargaining Preference Optimization (NBPO), Table 2 (primary cross-method evaluation)
Evaluation protocol: independent objective-wise win rate against the common reference,
judged by Llama-3.3-70B-Instruct on prompt-disjoint held-out prompts, both presentation orders.