COMPASS: Finding Where Reasoning Lives in Language Models

Pratyay Dutta; Kowshik Thopalli; Vivek Narayanaswamy

cs.AIOct 5, 2026

COMPASS: Finding Where Reasoning Lives in Language Models

Authors: Pratyay Dutta, Kowshik Thopalli, Vivek Narayanaswamy

Organizations: University of California, Riverside, CA 92521. Work performed during a summer internship at Lawrence Livermore National Laboratory · Lawrence Livermore National Laboratory, Livermore, CA 94550.

Abstract

Explicitly eliciting reasoning substantially improves LLM performance. Existing approaches require a predefined characterization of reasoning, whether through CoT prompt design, contrastive CoT directions, or via SAE derived reasoning features. For mathematical reasoning with verifiable answers, we show that a much simpler signal suffices, which is the correctness of the model's own direct answer attempts. This signal yields a latent direction that elicits reasoning. This direction is decodable within the activations of most attention heads, but only a small subset of them can be effectively intervened. We introduce COMPASS, an inference-time steering method that identifies these heads using a logit-space attribution score and steers their activations along the correctness direction, requiring only per-head activation statistics. Across three model families and multiple math benchmarks, COMPASS outperforms the activation-steering baselines we compare against, improves GSM8K accuracy by 16 percentage points on average, and approaches CoT accuracy with 20-70% fewer generated tokens. Interventions transfer without re-fitting to unseen benchmarks, and ablations show that both the correctness direction and the small set of heads carrying it are necessary, with the effect concentrated in remarkably few heads.

Figures & tables

Figure 1 : Left: The per-head correctness direction. In the activation space of a head, activations from correctly and incorrectly answered problems form two populations, with means $\mu_{+}^{l,h}$ and $\mu_{-}^{l,h}$ . $\theta_{l}^{h}$ is the unit vector connecting these means, and $\sigma_{l}^{h}$ is the standard deviation of the activations projected onto $\theta_{l}^{h}$ . Steering shifts an activation on the incorrect side by $\alpha\,\sigma_{l}^{h}$ along $\theta_{l}^{h}$ . Right: Head selection by attribution score. Top: the score $s_{l}^{h}$ of each head, with the $K$ heads selected within the promoted layer band outlined in blue. Bottom: the outcome contrast $\Delta p$ measured on the fitting rows. First generated tokens for correct answers concentrate on “To”, and for incorrect answers on “$”.

Panel of Figure 2 — Figure 2 : Steering elicits a short derivation, not a chain-of-thought transcript. The direct-answer prompt gets the answer wrong. With COMPASS, the same model reaches the correct answer with a short derivation, generating $3.0$ – $3.8\times$ lesser tokens than a chain-of-thought prompt.

	Llama-3.1-8B			Qwen3-4B			Gemma4-12B
	GSM8K	SVAMP	MATH-500	GSM8K	SVAMP	MATH-500	GSM8K	SVAMP	MATH-500
Standard Prompt	66.8 (93)	81.0 (48)	42.4 (345)	76.2 (68)	81.0 (22)	55.2 (250)	49.4 (67)	74.7 (3)	68.4 (347)
ITI Li et al. (2023)	73.8 (132)	68.7 (45)	23.6 (409)	71.9 (45)	44.3 (78)	30.6 (56)	22.3 (52)	64 (46)	30.3 (872)
Fractional Reasoning Liu et al. (2025)	79.9 (214)	81.0 (119)	42.2 (524)	88.3 (220)	89.3 (118)	77.8 (561)	65.4 (55)	85.6 (7)	72.7 (371)
COMPASS (ours)	83.7 (143)	85.7 (81)	45.2 (515)	89.8 (138)	90.7 (68)	77.8 (404)	88.0 (140)	79.3 (51)	79.8 (366)
CoT (oracle)	83.4 (205)	85.0 (157)	47.0 (788)	92.3 (303)	91.7 (224)	83.2 (924)	96.1 (338)	94.3 (261)	91.4 (730)

Table 1 : Accuracy (%) scores with mean generated tokens in gray. The heads selected by COMPASS lift the direct-answer (standard prompt) baseline toward the chain-of-thought and frontier oracles at a fraction of their token cost. ITI and Fractional Reasoning were re-run and we report their optimal performance.

	Llama-3.1-8B		Qwen3-4B
	GSM-Plus	HARP	GSM-Plus	HARP
Standard Prompt	60.4 (136)	21.5 (366)	65.6 (82)	35.4 (396)
COMPASS (ours)	71.1 (180)	22.7 (750)	80.4 (164)	49.1 (575)
CoT (oracle)	72.7 (247)	27.1 (1020)	84.1 (360)	56.9 (1505)
GPT-5.4-mini (oracle)	88.4 (385)	69.7 (1509)	88.4 (385)	69.7 (1509)

Table 2 : Transfer of steering heads and strength selected on GSM8K (to GSM-Plus) and MATH (to HARP), applied zero-shot, without re-tuning. Accuracy (%) with mean generated tokens in gray.

Figure 3 : Aggregated layer score The sum of the positive head-attribution scores in each layer. Yellow marks the layers from which the steering heads are selected based on the score (Eq 9 ).

Figure 4 : Best development set accuracy per candidate band. Bands are compared at constant

\alpha

and the comparison isolates which heads are steered rather than how hard. Yellow is the band the scan promoted (Llama 32–32 at

\alpha=15.66

; Qwen3 35–36 at

\alpha=25.53

). The dashed line is the unsteered direct-answer baseline on the same rows.

Figure 5 : Head-level ablations. Left: Head and direction controls, on the full 500-row split.

K

and

\alpha

are identical across bars. Only the intervention differs (error bars: sample standard deviation). Middle: Accuracy as only the

K

best-ranked heads are steered, with band and

\alpha

frozen. Right: Reverses the sign of the injected direction at the selected heads.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Dataset	Description
GSM8K [ 14 ]	Grade-school word problems with integer answers; 7,473 train and 1,319 test problems.
SVAMP [ 15 ]	1,000 one-step arithmetic word problems built by perturbing existing problems so that surface cues do not determine the operation; 700/300 train/test split of the public release (ChilleD/SVAMP). Train and test share no question text.
MATH-500 [ 34 ]	The 500-problem subset of the MATH test set [ 2 ] , seven subjects (prealgebra to precalculus) at five difficulty levels, LaTeX answers. No fitting is done on these 500 problems: the Llama selection is fitted on the 7,500-problem MATH train split, and the Qwen3 selection on the remaining 4,500 MATH test problems (MATH test minus MATH-500, matched on normalized problem text and verified disjoint).
GSM-Plus [ 16 ]	10,552 adversarial perturbations of GSM8K test problems across eight perturbation types. The 1,319 critical-thinking rows are unanswerable by construction (gold “None”) and are dropped, leaving 9,233 evaluated problems.
HARP [ 17 ]	4,780 short-answer problems from US national mathematics competitions (AMC, AIME, and others), six difficulty levels, LaTeX answers. No train split by design; steered with the MATH selection.

Appendix

Table 3 : Benchmarks. Train splits are fitting pools only; test splits are evaluated in full. GSM-Plus and HARP are used for transfer only.

Explore similar work

Math Reasoning in LLMs is Organized by Approach, Not Topic
Sep 22, 2026Sajad Goudarzi, Samaneh Zamanifard, Moloud Nasiri +1Mathematical Reasoning Benchmarks Large Language Model Tool Use
Clemson University
Soft Guidance Starts to Outperform CoT Prompting as LLMs Improve
Aug 4, 2026Denys Pushkin, Albert Q. Jiang, Aryo Lotfi +2Chain-of-Thought Reasoning Instruction-Tuned Models
EPFL · Apple · Mistral AI
Manifold-Guided Attention Steering
May 20, 2026Ian Li, Kapilesh Guruprasad, Raunak Sengupta +3Linear Activation Steering LLM Reasoning Strategies
University of California, San Diego

Sep 22, 2026, cs.AIJ/K move · Enter open · S save

Math Reasoning in LLMs is Organized by Approach, Not Topic

Sajad Goudarzi, Samaneh Zamanifard, Moloud Nasiri, Hamed Rahimian

Clemson University

Mathematical Reasoning Benchmarks Large Language Model Tool Use Text Corpora Virtual Cell

Open paper AskPDF arXiv

Mathematical reasoning benchmarks are typically organized by topic, but language models may organize their internal computation by reusable reasoning approach instead. In this paper, we investigate whether open math-capable LLMs organize internally by topical sub-skill or by reasoning approach, and we present evidence that the approach is the key. We introduce a generation-replay protocol: a model first generates a solution, after which we replay the exact prompt-plus-generation trajectory and extract activation-importance signatures over the reasoning tokens. We cluster these signatures without supervision across eight models and five mathematical reasoning sources, then evaluate the recovered structure with structural, semantic, and intervention tests. Across all 40 model-source cells, the recovered clusters outperform matched-size random baselines. Two independent frontier-LLM judges find approach-level coherence in 77-82% of real clusters versus 6-11% in within-source controls, and topic-pure clusters usually receive labels finer than the topic itself. In approach-controlled prompting, changing the requested reasoning approach shifts cluster assignment in seven of eight model conditions, whereas paraphrases largely preserve it. These results indicate that math-capable LLMs organize internal mathematical computation by reasoning approach rather than benchmark topic. The implication is that topic-stratified benchmarks and topic-balanced training corpora can still miss the axis that matters: even deliberately topic-balanced corpora may remain imbalanced over reasoning approaches.

Abstract

Explore similar work

Math Reasoning in LLMs is Organized by Approach, Not Topic

Soft Guidance Starts to Outperform CoT Prompting as LLMs Improve

Manifold-Guided Attention Steering