The Long Road to the Same Answer: Cognitive Bias Under Escalating Reasoning Budgets in Large Language Models
On the map: LLM reasoning → Reasoning benchmarks and evaluation also fits: Prompting and chain-of-thought reasoning
Abstract (arXiv)
Reasoning models allocate extra computation at inference time and present their answers as the product of deliberate thought. If this deliberation works the way dual-process accounts of human cognition suggest, longer thinking should weaken the classic decision biases that fast, intuitive judgment produces. Using 30 vignettes covering six biases (anchoring, framing, loss aversion, escalation of commitment, availability, confirmation) from an established benchmark, we run a dose-response study across four model families, pairing each reasoning model with a matched non-reasoning sibling and requesting thinking ceilings of 0, 1,024, 4,096, and 8,192 tokens, for 12,350 API calls. Because a requested ceiling is not the same as realized deliberation, we use the reasoning tokens each call consumed as the dose. First, reasoning models are not less biased than their siblings; the point estimate leans the other way in every family, but the item-level pooled contrast is not reliable (Delta = +0.031, t(29) = 1.45, p = .157). Second, bias magnitude does not reliably fall as realized deliberation grows: no slope is significantly negative, and where anything moves it is the signed score drifting further from the human direction. Third, anchoring is the only bias in the human direction (d = 1.89). Four of the other five lean the opposite way in all seven models; with five items per bias, that reversal is reliable for framing and directional for escalation of commitment, confirmation, and loss aversion, while availability is absent. A one-line instruction to restate the anchor before answering lowered anchoring on all five anchoring items, which no amount of additional thinking did, although the effect does not reach significance (p = .057). The results argue against treating test-time reasoning as a rationality guarantee and for auditing deployed models bias by bias.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 9 of its 25 references that are in the map. First posted on arXiv 7 Oct 2026.
Builds on
9 of 25 references.
Closest ancestors · 1
Same direction, newest first
- 2025Qwen3 Technical Report14 May
Other references · 8
- 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
- 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,928
- 2025s1: Simple test-time scalingcited by 1,311
- 2024OpenAI o1 System Cardcited by 1,220
- 2025Reasoning Models Don't Always Say What They Thinkcited by 388
- 2025The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexitycited by 298
- 2024A Comprehensive Evaluation of Cognitive Biases in LLMscited by 32
- 2025The Bias is in the Details: An Assessment of Cognitive Bias in LLMscited by 10
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026Reasoning aligns language models to human cognition
- 2026To Think or Not To Think, That is The Question for Large Reasoning Models in Theory of Mind Tasks
- 2023A Systematic Comparison of Syllogistic Reasoning in Humans and Language Models
- 2026Does Thinking Help Fairness? Reasoning Tokens Resolve Some Biases but Create More
- 2026Inference-Time Reasoning Selectively Reduces Implicit Social Bias in Large Language Models
- 2024Comparing Inferential Strategies of Humans and Large Language Models in Deductive Reasoning
- 2025Unveiling Confirmation Bias in Chain-of-Thought Reasoning
- 2026BiasTrace: Linking Reasoning Behaviours to Biased Outputs in LLMs