# How Much Does a Reasoning Summary Reveal? An Observability Ladder for Large Language Models

arXiv: https://arxiv.org/abs/2608.02089 · HTML: https://mlsanity.com/paper/2608.02089/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 22 of its 57 references that are in the map.
- First posted on arXiv 3 Aug 2026.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 22 of 22)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Language Models (Mostly) Know What They Know](https://mlsanity.com/paper/2207.05221/) (2022)
- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [Representation Engineering: A Top-Down Approach to AI Transparency](https://mlsanity.com/paper/2310.01405/) (2023)
- [Measuring Faithfulness in Chain-of-Thought Reasoning](https://mlsanity.com/paper/2307.13702/) (2023)
- [Reasoning Models Don't Always Say What They Think](https://mlsanity.com/paper/2505.05410/) (2025)
- [Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation](https://mlsanity.com/paper/2503.11926/) (2025)
- [Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety](https://mlsanity.com/paper/2507.11473/) (2025)
- [Efficient Few-Shot Learning Without Prompts](https://mlsanity.com/paper/2209.11055/) (2022)
- [Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought](https://mlsanity.com/paper/2603.05488/) (2026)
- [Can LLMs Predict Their Own Failures? Self-Awareness via Internal Circuits](https://mlsanity.com/paper/2512.20578/) (2025)
- [A Pragmatic Way to Measure Chain-of-Thought Monitorability](https://mlsanity.com/paper/2510.23966/) (2025)
- [Benchmarks Saturate When The Model Gets Smarter Than The Judge](https://mlsanity.com/paper/2601.19532/) (2026)
- [Lexical Hints of Accuracy in LLM Reasoning Chains](https://mlsanity.com/paper/2508.15842/) (2025)
- [How to Steal Reasoning Without Reasoning Traces](https://mlsanity.com/paper/2603.07267/) (2026)
- [MonitorBench: A Comprehensive Benchmark for Chain-of-Thought Monitorability in Large Language Models](https://mlsanity.com/paper/2603.28590/) (2026)
- [Catching rationalization in the act: detecting motivated reasoning before and after CoT via activation probing](https://mlsanity.com/paper/2603.17199/) (2026)
- [Measuring Weak-to-Strong Legibility of Reasoning Models](https://mlsanity.com/paper/2603.20508/) (2026)
- [ReasonOps: Operator Segmentation for LLM Reasoning Traces](https://mlsanity.com/paper/2605.29192/) (2026)
- [Length Penalties Make Chain-of-Thought Less Monitorable](https://mlsanity.com/paper/2607.09786/) (2026)

## Built on by (most cited first, 2 of 2)

- [The Answer Is Not the Argument](https://mlsanity.com/paper/2609.00264/) (2026)
- [Hallucination Across the Reasoning Lifecycle: Interface Visibility, Causal Evidence, and Release Control in Large Reasoning Models](https://mlsanity.com/paper/2610.05472/) (2026)

## Abstract (arXiv)

Large language models often show users a final response and a short reasoning summary while the full reasoning trace stays hidden. We introduce an observability ladder that holds each completed run fixed and varies only what a reader inspects to judge whether the answer is correct: the response, a self-summary the model writes from the trace, the trace itself, and internal signals, each with and without the prompt. Across three benchmarks and five open-weight Qwen3 and gpt-oss models, we train matched linear correctness predictors on each access level. Without the prompt, summaries carry most of the trace's ranking signal (mean AUROC 0.774 versus 0.813) and add +0.156 over the response alone. With the prompt visible, the summary's gain collapses to +0.019, while the trace still adds +0.041. Even at equal length, the trace's last words predict correctness as well as summaries, or slightly better, and carry denser and more discriminative uncertainty and self-correction cues. On MMLU-Pro questions with both correct and incorrect runs, linear summary readers are near chance and trace readers retain only modest signal, both with and without the prompt (prompt-withheld AUROC 0.503-0.545 versus 0.544-0.590). With the prompt withheld, a GPT-5-mini reader recovers substantially more signal from both summaries and traces on gpt-oss-20b, and even then the trace keeps a small +0.034 advantage. Much of the linear readers' trace signal is associated with length. In the common case where users already hold the prompt, summaries are less helpful than the full trace for monitoring correctness. Monitorability is thus a joint property of the display and the reader, so any monitorability claim, including for faithfulness, should specify both.