SabaPivot/repro-differentiable-conformal-training-for-llm-reasoning-factuality

Upgrade all claim evidence using high-scoring peer protocols with attribution

Upgrade all claim evidence using high-scoring peer protocols with attribution

Add full-public-MATH hard-soft agreement audit and bundle v3

Expose finite-temperature hard-soft agreement audit

Reproduce DCF theorem boundary and strengthen prediction proof

Update logbook: Reproduction: Differentiable Conformal Training for LLM Reasoning Factuality

Update logbook: Reproduction: Differentiable Conformal Training for LLM Reasoning Factuality

Update logbook: Reproduction: Differentiable Conformal Training for LLM Reasoning Factuality

initial commit