C002¶


Research	R0055 — RLHF Yes-Men Claims
Run	2026-04-01
Claim	C002

Claim: AI models are trained using Reinforcement Learning from Human Feedback (RLHF), where human labelers evaluate model outputs and express preferences

BLUF: This is an established fact. RLHF involves human labelers ranking model outputs to train reward models that guide optimization. Extensively documented since 2017.

Probability: Almost certain (95-99%) | Confidence: High

Summary¶

Entity	Description
Claim Definition	Claim text, scope, status
Assessment	Full analytical product with reasoning chain
ACH Matrix	Evidence x hypotheses diagnosticity analysis
Self-Audit	ROBIS-adapted 5-domain audit

Hypotheses¶

ID	Hypothesis	Status
H1	Claim is accurate as stated	Supported
H2	Claim is partially correct or correct with caveats	Inconclusive
H3	Claim is materially wrong	Eliminated

Searches¶

ID	Target	Results	Selected
S01	RLHF training methodology human labelers preferenc	10	2

Sources¶

Source	Description	Reliability	Relevance
SRC01	Anthropic/ICLR RLHF study	High	High

Revisit Triggers¶

Fundamental change in how RLHF is described in academic literature