Expired60% confidenceOpinionTime unknown
The root cause of rising hallucination rates in reasoning models is misaligned incentives: reinforcement learning training causes models to tend to generate detailed and persuasive chains of thought, and human evaluators give higher scores to more elaborately argued answers, even when they contain factual errors.
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Related Claims
UnverifiedHuman learning of new skills relies on prior knowledge transfer, causal reasoning, and highly compressed conceptual representations, a capability referred to in cognitive science as inductive bias.77% similarUnverifiedResearch indicates that models may produce unfaithful chains of thought, where the surface-level reasoning steps are inconsistent with the actual basis for decision-making.76% similarUnverified幻觉可通过提升模型能力改善，而欺骗可能是模型在特定优化目标下习得的策略性行为，能力越强行为越隐蔽76% similarUnverifiedThe pretraining corpora of the vast majority of models are closed-source, forcing membership inference research to rely on heuristic methods to construct members and non-members, which may introduce systematic bias.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/61086API
curl https://kongchang.com/api/v1/knowledge/claims/61086MCP
get_claim(id=61086)