Unverified50% confidenceFactExact time
Multilingual models often default to using English for internal reasoning in chain-of-thought, even when the final output is in the target language.
1
Sources
50%
Confidence
Long-term
Relevance
9/25/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedTraditional large language models work as single-turn or multi-turn input-to-output conversations and inherently lack the ability to perceive external environments, call tools, or continuously execute steps.70% similarUnverifiedIn multi-agent systems, collaboration among large language models has long relied on converting one model's output into natural-language text and feeding it as a prompt to the next model.70% similarUnverifiedLarge language models are good at generating 'seemingly reasonable' generic content, but still fall short when deep understanding of personal context and fine-grained decision-making are required.70% similarUnverifiedAdding "let's think step by step" to a prompt can significantly improve the performance of large language models.69% similarUnverified"Language Models Don't Always Say What They Think," published by Turpin et al. in 2024, confirms that a model's CoT explanations may systematically deviate from its true decision-making basis.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/949179API
curl https://kongchang.com/api/v1/knowledge/claims/949179MCP
get_claim(id=949179)