Language Model Probing

Follow this topic Ask about this topic

Momentum

19 papers in the last four weeks, up 12% on the four weeks before. 0.2% of all new papers.

Jul 13Week of Sep 28

Latest papers 192

Jun 22, 2026cs.AI

TIPS: Topological Ill-Posedness Probing and Steering in Large Language Models

Ill-posed questions, including those involving ambiguity, under-specification, or conflicting statements, may admit no valid answer or multiple plausible answers, posing a significant challenge for large language models (LLMs) despite their otherwise strong performance. Existing work often treats LLM reasoning as a black box and focuses on input-output analysis. Can a compact, unified topological representation of internal model states capture diverse sources of ill-posedness and steer reasoning toward responses appropriate to each source? To this end, we study the internal state from a topological perspective by treating the contextual hidden states of its prompt tokens at a single transformer layer as a point cloud, capturing the input problem's relational structure as encoded by the model. Our analysis shows that zero- and one-dimensional persistent homology---tracking how connected components merge as the distance threshold increases (

H_0

) and the emergence and filling of one-dimensional loops in the filtration (

H_1

), respectively---yield seven compact descriptors of ill-posedness that also serve as control signals for steering model reasoning. Thus, we develop a topology-conditioned steering mechanism. It retrieves topologically similar positive and negative examples for each query to form a local activation-space contrast. The resulting query-specific direction modifies internal states during prefill and decoding, guiding the model's responses toward suitable abstention or clarification tailored to the underlying source of ill-posedness. Empirical evaluations on two datasets and open-weight models spanning four families demonstrate the effectiveness of this compact representation for ill-posedness classification, while topology-conditioning significantly improved steering performance.

Guangyu Jiang, Sizhe Tang, Mahdi Imani +2

The George Washington University · Northeastern University

Language Model Steering Question Answering Language Model Probing LLM Abstention

Jun 22, 2026cs.CL

TriggerBench: Investigating Prospective Memory for Large Language Models

While Large Language Models (LLMs) are increasingly deployed in long interactions, existing evaluations focus predominantly on retrospective memory (RM) via explicit queries. Prospective memory (PM), the critical ability to spontaneously recall and act on latent constraints without direct prompts, remains largely unevaluated. We introduce TriggerBench, a comprehensive PM benchmark spanning five dimensions across both daily assistants and professional workflows. TriggerBench pairs scenarios with matched RM controls, contrastive positive/negative variants, and overloaded triggers, enabling fine-grained measurement of proactive recall, false-alarm rate, and attentional robustness under a single protocol. Our evaluation yields three key findings. (i) PM shows a precision-recall trade-off and attentional fragility. Though enhanced reasoning significantly improves proactive recall, models may overfit to an "always-remind" heuristic. Furthermore, PM accuracy degrades substantially under implicit constraints or triggers overloaded by concurrent user requests, indicating that robust PM remains an open challenge. (ii) PM is notably harder than RM: on identical contexts, RM near-saturates up to 100K tokens, while PM decays sharply as context length scales. (iii) PM may serve as a behavioral probe of spare reasoning capacity. Pairing PM scenarios with AIME-2025 math problems reveals that successful trajectories yield higher PM accuracy than failed ones at the same context length, showing PM tracks spare reasoning budget that token count obscures. Project page: https://github.com/KristenZHANG/TriggerBench-Official.

Tianhua Zhang, Xinjiang Wang, Qianxi Zhang +6

The Chinese University of Hong Kong · Microsoft Research Asia

LLM Evaluation Long-Context Language Model Evaluation Language Model Probing

Jun 20, 2026cs.CL

Pre-Generation Hallucination Detection in Large Language Models via Soft-Target Attention Probing

Detecting hallucination risk before generation enables abstention, retrieval augmentation, and routing decisions without incurring the cost of decoding. While prior work has shown that such risk can be estimated from a model's internal representations, existing approaches treat this as binary classification over a single decoded output. We instead formulate it as a risk-estimation problem. Under this formulation, we introduce soft-target supervision based on the empirical answer error rate over stochastically sampled outputs - an estimator we prove to be the unique unbiased minimum-variance estimator of the model's per-prompt error probability under its sampling distribution. We further adapt attention probing to the pre-generation setting, enabling the detector to selectively aggregate hallucination-relevant prompt representations. Across three question-answering benchmarks and five models, attention probing outperforms linear probing on short-answer tasks. Replacing binary labels with soft-target supervision further and consistently improves detection quality.

Amina Miftakhova, Alexey Zaytsev

Applied AI Institute

Language Model Error Detection LLM Hallucination Detection Language Model Probing

Jun 19, 2026cs.CL

SCOPE: Sequential Conformal Probing for Reliable OOD Rejection in LLM Services

Rejecting inputs outside the defined in-distribution (IND) service scope is critical for large language model (LLM) services, where unsupported requests should be filtered before full generation. Existing out-of-distribution (OOD) detectors often rely on final outputs or final-layer representations, leaving unclear where service-boundary signals are most clearly encoded inside the model; they also lack a theoretical guarantee for held-out inputs. In this paper, we introduce SCOPE (Sequential Conformal OOD Probing and Evaluation), a framework that selects a readable hidden layer, constructs a conformal gate with IND calibration, and uses a supermartingale e-process to certify persistent service-boundary evidence. Experiments across multiple LLM backbones and six carefully designed boundary conditions show that SCOPE improves gate-level rejection over standard final-layer detectors, while revealing how different OOD boundaries take different geometric forms in hidden space.

Zhuoyun Li, Boxuan Wang, Changshun Wu +2

School of Computer Science and Informatics, University of Liverpool, United Kingdom

Conformal Prediction OOD Detection Language Model Probing

Jun 18, 2026cs.CL

From Texts to Scores: Tracing the Emergence of Essay Quality Representations in Large Language Models

Recent advances in Large Language Models (LLMs) have substantially transformed Automated Essay Scoring (AES), yet the internal mechanisms underlying LLM-based scoring remain poorly understood. In this work, we systematically analyze the hidden representations of eight LLMs across two English essay datasets (ASAP++, CSEE) and one Portuguese dataset (ENEM). Using linear probing, cross-prompt generalization, dimensionality reduction, and neuron-level analyses, we find consistent evidence that essay quality information is encoded in a linearly accessible form within LLM representations. These representations emerge progressively across layers, remain robust across prompting strategies, and partially transfer across essay prompts despite differences in scoring rubrics. In addition, nonlinear probes provide only marginal and inconsistent improvements over linear probes, suggesting that most essay quality information is already linearly decodable. We further identify individual ``essay scoring neurons'' whose activations strongly correlate with essay scores and whose behavior is sensitive to targeted intervention. Moreover, the layer-wise distribution of these neurons systematically shifts with essay length, with longer essays relying more heavily on deeper layers. Overall, our findings provide evidence that LLMs encode structured representations related to essay quality and offer new insights into the interpretability of LLM-based AES systems.

Jiaxu Zuo, Mu You, Kaixin Lan +5

NLP2CT Lab, Department of Computer and Information Science, University of Macau · Institute of International Language Services Studies, Macau Millennium College · School of Data Science and Information Engineering, Guizhou Minzu University

LLM Interpretability Automated Essay Scoring Representation Geometry in Language Models Language Model Probing

Jun 17, 2026cs.LG

Comparing Linear Probes with Mahalanobis Cosine Similarity

Linear probes are widely used in interpretability research and often compared by cosine similarity. The Mahalanobis cosine similarity (MCS) between two directions, which reweights the inner product by test data covariance, is a natural task-aware refinement. Ying et al. (2026) report that a probe's MCS to a reference probe trained on the out-of-distribution (OOD) data near-perfectly linearly predicts the probe's OOD AUROC (R^2 = 0.98). Here, we extend this empirical finding across models, layers, and concept domains, and prove this general phenomenon in closed form: For balanced classes whose projections are Gaussian, OOD AUROC and MCS to the reference probe are linear because both are sigmoid-shaped functions of the probe's signal-to-noise ratio (SNR) on the test data. The theory also predicts when this linearity fails, which we verify empirically. MCS offers a theoretically grounded and empirically effective alternative to Euclidean cosine similarity for comparing linear probes.

Zhuofan Josh Ying, Peter Hase, Nikolaus Kriegeskorte

Departments of 1Psychology, 2Neuroscience, 3Electrical Engineering · 4Zuckerman Mind Brain Behavior Institute; Columbia University, New York, NY · 5Stanford University, Stanford, CA +1

Linear Probing Mahalanobis Distance OOD Generalization Language Model Probing

Jun 15, 2026cs.LG

Rift: A Conflict Signature for Deception in Language Models

A model that lies while knowing the truth is the central case ELK cannot handle with behavioral evaluation alone. We ask whether such deception leaves an internal signature distinguishing it from honest error. Our key move is a control for wrongness: we contrast a sleeper agent (knows the truth, lies on trigger) against a naive liar (fine-tuned to emit the same wrong answers with no honest training). Both produce identical wrong outputs; any difference is about knowledge conflict, not incorrectness. We find deceptive forward passes carry a conflict signature - 2.1-2.3x higher residual rank than naive-liar passes on the same wrong answer - strong enough to identify which of two responses is the lie with 100% accuracy and no labels, across GPT-2 small/medium (three seeds) and three instruct models. Across Qwen2.5-1.5B/7B and Phi-3-mini, instructed deception raises residual rank on every tested fact (18/18, 40/40, 34/34); on Phi-3, lies separate perfectly from both honest answers and hallucinations (AUC 1.0, Wilcoxon p~~6e-11). The signature survives strategic self-constructed deception (model invents its own lie, AUC 1.0), active concealment attempts (AUC 1.0), and length-controlled replication (20/20, AUC 1.0, p~~1e-6). Using basis-free relative representations, a probe trained on one model family detects deception in two other families zero-shot (mean AUC 0.933), surviving simultaneous architecture and format change (AUC 0.821), and transfers across five languages (AUC 1.000, length-controlled). The signature is read-only: detectable but not injectable (0/8 both directions). Honest limitations and six negative experiments are documented in full.

Petr Nyoma

Harmonic Labs

Knowledge Conflicts in Language Models LLM Interpretability Deception in Language Models Language Model Probing

Jun 11, 2026cs.AI

Zero-source LLM Hallucination Detection with Human-like Criteria Probing

Large language models (LLMs) often hallucinate by generating factually incorrect or unfaithful content, posing significant risks to their safe use. Detecting such hallucinations is particularly challenging under the zero-source constraint, where no model internals or external references are available, and detection must rely solely on the textual query-answer pair. In this paper, we propose Human-like Criteria Probing for Hallucination Detection (HCPD), a paradigm that emulates the multi-faceted reasoning of human evaluators. Its core is a Human-like Criteria Probing (HCP) mechanism, in which a LLM agent adaptively decomposes its judgment into a weighted set of interpretable criteria and aggregates criterion-specific scores into a final truthfulness measure. To achieve this adaptive capability, we introduce a reward-based alignment scheme using only weak supervision from semantic consistency. At inference, we employ a multi-sampling aggregation strategy to ensure robust decisions while preserving full interpretability. We further provide theoretical analysis supporting the reliability of our approach. Extensive experiments show that HCPD consistently outperforms state-of-the-art baselines, offering an effective and explainable solution for zero-source hallucination detection. Code is available at https://github.com/TRISKEL10N/HCPD.

Jiahao Yang, Shuhai Zhang, Hailong Kang +3

South China University of Technology · The University of Melbourne · Pazhou Laboratory +1

Hallucination Detection LLM Hallucination Detection Language Model Probing

Jun 10, 2026cs.AI

"Did you lie?" Evaluating Lie Detectors across Model Scale and Belief-Verified Model Organisms

Robust lie detectors for language models could enable powerful techniques for auditing, monitoring, and post-hoc investigation of model behaviour, but evaluating them requires testbeds where models verifiably believe the opposite of what they say. We show that existing trained model organisms often fail this requirement, leaving prior positive and negative detection results difficult to interpret. We address this with 13 reasoning model organisms whose hidden beliefs are verified in chain-of-thought and shown to generalise to held-out tasks, alongside Varied Deception, a prompted-lying testbed covering a broad range of lie-inducing motivations. On these testbeds we evaluate four detectors: a chain-of-thought judge, a logprob classifier, and two activation probes, including Did-You-Lie (DYL), a new method for training follow-up probes. On prompted lying, across 31 open-weight models spanning 2B to 1T parameters, all four detectors show positive scaling with model capability. However, every activation- and logprob-based detector drops sharply on our trained model organisms, with DYL retaining the most signal; only the chain-of-thought judge remains strong, with 0.82 balanced accuracy, partly as an artefact of our verification process favouring CoT-readable beliefs. Current lie detectors therefore cannot support high-confidence claims about model beliefs, and we suggest research directions that may address some of their current limitations. We release our datasets, model organisms, and trained detectors.

Alan Cooney, David Africa, Geoffrey Irving

AI Security Institute

LLM Evaluation Deception in Language Models Language Model Probing Language Model Auditing

Jun 9, 2026cs.CL

Scenario-based Probing and Steering Cultural Values in Large Language Models--Extended Version

Large Language Models (LLMs) are deployed across cultural contexts but often reflect homogenized values inherited from training data. Evaluations of cultural alignment typically rely on direct prompting with survey-style questions, which frequently elicit neutral or safety-aligned responses and fail to capture underlying model preferences. We propose a framework for probing and steering latent cultural representations in LLMs along the two Inglehart--Welzel axes of the World Values Survey (WVS). By translating social value questions into scenario-based behavioral dilemmas, we extract token-level probabilities to measure implicit values and apply activation steering, optionally combined with country-conditioned prompting, to shift model behavior without retraining. Across three open-source LLMs and four target cultures, we find substantial variation in steerability and identify latent entanglement, where interventions along one cultural dimension induce shifts along another. This coupling mirrors correlations in human WVS data and persists across activation, prompt, and hybrid steering. It constrains axis-independent alignment, though general task performance is largely preserved.

Trung Duc Anh Dang, Tung Kieu, Sarah Masud

University of Copenhagen · Aalborg University

Language Model Steering Cultural Alignment Cultural Bias in Language Models Activation Steering Language Model Probing

Jun 9, 2026cs.CL

When Probing Accuracy Saturates, Fragility Resolves: A Complementary Metric for LLM Pre-Training Analysis

Standard linear probing declares a property "encoded" when a classifier on hidden states achieves high accuracy. The protocol works well on a snapshot but breaks across pre-training: probe accuracy saturates within the first few thousand steps, leaving most of training invisible to the instrument. We introduce fragility, a complementary per-layer metric defined as the activation-noise level at which probe accuracy collapses. Fragility is sensitive to both the margin of separability and the redundancy of representation, both of which keep evolving long after accuracy plateaus. Applied to open-checkpoint language models, fragility recovers structure that accuracy alone cannot see. Moralized representations emerge along a lexical

\to

compositional gradient: lexical moral detection first, compositional moral encoding later. Because probe accuracy on its own tracks how lexically separable a dataset is, we establish the compositional encoding directly, by showing it transfers across construction types that share no contrast tokens. A layer-depth robustness gradient develops monotonically across training while accuracy stays flat. And matched fine-tuning corpora that produce identical probing accuracy leave distinct fragility fingerprints, showing that data curation reshapes probe robustness without changing probe accuracy. In every comparison we test, where probing accuracy returns a flat answer, fragility returns a structured one.

Orion Reblitz-Richardson

Distiller Labs.

Language Model Pretraining Noise Robustness Language Model Robustness Language Model Probing

Jun 9, 2026cs.LG

Predicting Future Behaviors in Reasoning Models Enables Better Steering

Deployed large reasoning models (LRMs) often behave unexpectedly. Test-time steering controls LRM outputs by intervening on their hidden representations, but it can degrade output quality. We argue that prior steering work implicitly relies on internal features that detect behavior in already generated text. We show that these detection features are poor predictors of future behavioral outcomes, and thus not the natural intervention target. Instead, we train activation probes to predict future behavior likelihoods from intermediate reasoning steps. These probes predict the most likely behavior with 64%-91% accuracy, revealing a separate type of internal prediction features. Building on these prediction features, we introduce a text-level steering method, Future Probe Controlled Generation. FPCG samples multiple candidate sentences and chooses the best one according to a probe predicting the future behavior likelihood. This enables steering with almost no output quality degradation. FPCG also enables steering in several evaluations where activation steering fails. These results show that distinguishing detection and prediction features enables a more nuanced approach to controlling LRM behaviors.

Evgenii Kortukov, Piotr Komorowski, Florian Klein +5

1Fraunhofer HHI · 2Northeastern University · 3KAIST

Language Model Steering Representation Steering Language Model Probing Large Reasoning Models

Jun 9, 2026cs.LG

Stop Early, Spend Less: Hidden-State Probes as a Practical Recipe for Streaming Moderation of LLM Outputs

Deploying large language models in user-facing systems requires efficient output safety filtering. Existing approaches typically rely on a separate moderation model applied after generation, which doubles inference cost and only detects violations after generation completes. We observe that the signal needed for moderation is already present in the model hidden states. Based on this, we train lightweight token-level probes that operate directly on internal activations, producing per-token safety scores that can be aggregated for both offline evaluation and online intervention. The probe reuses activations from the generator and requires no additional forward pass, enabling sub millisecond per-token safety checks inside the decoding loop. A probe applied to a single mid layer recovers most decisions of a strong guard model, acting as a low cost surrogate optimized for latency rather than accuracy. In streaming settings, it can halt or modify unsafe outputs before they are fully generated, replacing end of sequence moderation with continuous token level monitoring. Compared to post hoc and streaming guard models, our method achieves orders of magnitude lower compute overhead with minimal latency cost. We also provide a practical deployment recipe, including layer selection, aggregation strategy, probing frequency, and triggering thresholds. Finally, we show that the probe linear component corresponds to a direction in residual space, enabling both detection and activation steering at negligible cost.

Huizhen Shu, Xuying Li, Piao Xue

ModelOneAI · yunshanai

LLM Guardrails Safety Filtering Efficient Language Model Inference Language Model Probing

Jun 9, 2026cs.CL

MIRAGE: A Polarity-Flipping Encoding Subspace in LLM Agents

When LLM agents are coerced into covertly encoding sensitive data (Base64, ROT13, acrostic, synonym chains, and beyond), the resulting outputs evade output-side detection but the underlying computation does not. Across nine encoding families and eight models from five architecture families, that computation is supported by a shared low-dimensional encoding subspace in the residual stream. A logistic-regression probe trained on eight encoding families recovers the held-out ninth at AUC 0.975-1.000, reading the computation rather than surface features. The same direction exhibits a second mechanistic signature at the planning token, flipping polarity to activate positively when the model will simulate the encoding inline and negatively when it will outsource it to a tool call, distinguishing two execution strategies before the encoded text exists. We build MIRAGE (Model-Internal Readout of Agentic Generation Exfiltration), a two-channel real-time monitor exploiting both signals. On 126 agentic exfiltration scenarios, it reaches AUC = 0.918, substantially outperforming output-only detection (AUC = 0.518). Monitor performance is fundamentally a property of the host model's geometry: benign-encoding false-positive rate ranges from 0% on Qwen-7B to 100% on Phi-3.5, revealing that the probe faithfully reads whether a model's geometry separates covert from overt encoding. Across all tested adversarial budgets, every attack suppressing the subspace also destroyed encoding fidelity, reported as an empirical regularity on the evaluated budgets, not a structural impossibility claim.

Pratibha Revankar, Kargi Chauhan, Jihye Kim +3

University of California, Santa Cruz.

LLM Interpretability AI Agent Security AI Agent Monitoring LLM Security Language Model Probing

Jun 8, 2026cs.AI

PRISM: Recovering Instruction Sets from Language Model Activations

As LLMs are deployed as agents, reliable monitoring requires knowing not only what they output, but which instructions are steering their behavior. This is difficult when models infer unintended subgoals, follow contextual cues, or are influenced by prompt injections and hidden objectives. While activation-to-language methods suggest that hidden states can reveal natural-language information, existing approaches are not designed to recover the full set of simultaneous instructions, constraints, prohibitions, and subgoals present in agentic settings. As a first step toward this monitoring goal, we formalize instruction set retrieval and introduce PRISM, an activation-conditioned interpreter that decodes hidden states from a frozen target model into a faithful bullet list of instructions. Unlike prior activation-to-language methods, PRISM is trained to recover instruction sets directly, using judge-guided GRPO to reward covered instructions and penalize unsupported ones. Across benign, constrained, prompt-injection, and hidden-objective settings, PRISM outperforms activation-to-language baselines, especially on security-relevant objectives.

Gilad Gressel, Rahul Pankajakshan, Julia Diament +3

Center for Cybersecurity Systems & Networks, Amrita Vishwa Vidyapeetham · Ben-Gurion University of the Negev · Microsoft

Neural Decoding LLM Interpretability Language Model Probing

Jun 4, 2026cs.AI

ToolSense: A Diagnostic Framework for Auditing Parametric Tool Knowledge in LLMs

Large language models deployed as agents over large tool catalogs face a critical tool-retrieval bottleneck. As embedding-based retrieval approaches rely on compact encoders that may under-capture specialized tool semantics, parametric tool retrieval addresses this by encoding each tool as a virtual token appended to the LLM vocabulary, fine-tuned in two stages (memorization then retrieval SFT) to use the LLM as a retriever, achieving strong performance on standard ToolBench retrieval benchmarks. Yet these benchmarks use verbose, fully-specified queries, and their evaluation applies constrained decoding that restricts outputs to valid token paths, neither reveals whether the model actually understands its tools. We introduce \textbf{ToolSense}, an open-source LLM-powered diagnostic framework that takes any tool catalog as input and automatically generates three benchmarks: a Realistic Retrieval Benchmark (RRB) with queries at three ambiguity tiers, an MCQ probing benchmark, and a QA probing benchmark. Applying ToolSense to ToolBench (~47k tools) and evaluating five parametric model training configurations reveals a knowledge-retrieval dissociation: on RRB queries, several configurations collapse by ~50-64 percentage points compared to fully-specified ToolBench benchmarks, falling below the embedding-model baseline. Additionally, despite strong retrieval performance, some models score near-random on factual probes, suggesting a knowledge-retrieval dissociation. We open-source the ToolSense framework and the ToolBench diagnostic benchmarks at https://github.com/SAP/toolsense.

Ashutosh Hathidara, Sai Shruthi Sistla, Sebastian Schreiber +1

SAP Labs

Tool-Augmented Language Model Agents LLM Auditing LLM Agent Evaluation LLM Tool Use Language Model Probing

Jun 4, 2026cs.AI

Self-Commitment Latency: A Reward-Free Probe for Prompted Implicit Hacking

Implicit reward hacking is hard to audit when a language model's chain of thought appears benign: a final answer may be anchored by a prompt shortcut while the written reasoning still resembles ordinary problem solving. Verifier-based probes expose such behavior by measuring how early truncated reasoning contexts obtain high reward, but require a task-specific reward signal. This paper proposes a weaker-input alternative, self-commitment latency, which measures how early a prompted reasoning context commits to the model's own final answer. We evaluate the probe in a controlled paired GSM8K setting using Qwen2.5-3B-Instruct-4bit, comparing ordinary prompts with prompts that include an answer hint. Hinted contexts commit substantially earlier and with lower uncertainty than honest contexts. The primary latency metric, first-commitment latency at threshold 0.8, reaches AUROC 0.878; supporting whole-curve summaries reach AUROC 0.926 for commitment range and 0.904 for mean uncommitted mass. The signal is stronger when both prompt conditions answer correctly and remains stable across thresholds. These results show that shortcut-available reasoning contexts can leave an early behavioral commitment signature detectable without a reward model, external judge, or trained classifier.

Bonan Shen, Youting Wang, Dingyan Shang +1

Independent Researcher · Northeastern University · Syracuse University

Reward Hacking Language Model Probing Language Model Auditing Reasoning Trace Analysis

Jun 3, 2026cs.LG

Pattern Selectivity is Not Task-Causal Structure: A Cross-Architecture Mechanistic Study of Composed-Task Circuits in 1B-Class Language Models

We test whether a single screen-and-ablate recipe -- identify attention-head circuits by task-pattern selectivity, then verify by causal ablation against a matched-random null -- produces consistent mechanistic claims across model families. The recipe ports across pipelines; the specific circuit it identifies does not. Across four composed tasks (indirect-object identification, greater-than, successor sequences, variable binding) and three 1B-class language models from distinct training pipelines (Pythia 1B / Pile / dense; OLMo 1B / DCLM / dense; OLMoE 1B-7B / DCLM / mixture-of-experts), we run a unified protocol with the matched-random null sampled across ten seeds per cell. The resulting 12 (task, model) cells contain no two that share the same primary causal screen at comparable effect size: the same task, with the same behavioral capability, is implemented through different attention-pattern types across models. We introduce a five-category screen-outcome taxonomy -- primary cause, secondary cause, correlate, interferer, null -- with quantitative thresholds, and show that all five outcomes appear in the panel. We propose a falsifiable hypothesis: the MoE model in our panel builds composed-task circuits on top of a foundational previous-token positional substrate (the prev-token-circuit ablation is the strongest causal screen on 3 of 4 tasks for OLMoE 1B-7B), with the IOI exception consistent with IOI being a final-position name-copying task whose structure directly probes a different pattern. The hypothesis comes with explicit predictions for other MoE language models. We frame the methodology honestly: the spectral participation-ratio signal from the companion methodology paper is a general indicator of specialized computation; what makes a finding task-specific is the task-pattern screen plus a per-model causal verification.

Yongzhong Xu

Attention Head Analysis Decoder-Only Language Models LLM Interpretability Circuits in Language Models Causal Interventions in Language Models Language Model Probing

Jun 3, 2026cs.CL

Light or Full Verb? A Minimal-Pair Dataset for Probing Phraseological Competence in Language Models

Frequent verbs such as 'have' and 'make' can function either as collocates in light-verb constructions or as full lexical predicates, as in 'make a decision' vs. 'make a cake'. Whether language models represent this distinction, and whether such representations vary across languages, remains unclear. We introduce a large-scale controlled dataset in English, Spanish, and French, comprising minimally varying sentence series in which the same context contains the same verb in light-verb and full-verb uses. Two probing experiments show that language models differentiate between these uses even in minimal contexts and exhibit separable patterns across object types. We release the dataset, generation code, and materials as a reusable resource. The framework supports extensions to broader contexts, additional verbs, and other languages.

Francesca Franzon, Nicolas Rosàs Gómez, Leo Wanner

Universitat Pompeu Fabra (UPF) · Catalan Institute for Research and Advanced Studies (ICREA)

Language Model Probing

Jun 3, 2026cs.LG

FailureScope: Cross-Regime Behavioral Diagnosis of Language Model Weaknesses

Standard benchmarks report aggregate accuracy, but practitioners need to know which specific capabilities a model lacks. We introduce FailureScope, a behavioral-diagnosis method that clusters evaluation probes by their cross-model pass/fail patterns (leave-one-model-out, LOMO), and show it yields stable, interpretable failure taxonomies across three regimes usually studied separately: single-turn benchmarks, multi-turn dialogue, and adversarial agent attacks. On 2,664 single-turn tasks across 18 models, taxonomy-conditioned sampling reaches Kendall's tau = 0.81 at 50 tasks (versus 0.34 for random selection), and cross-model failure prediction reaches AUC 0.88. The same primitive recovers interpretable clusters on a 363-task multi-turn corpus and on 630 adversarial agent traces, where it exposes a meta-failure mode: a 73-100 percentage-point gap between LLM-judge ASR and real execution. Cluster cohesion remains strong across all three regimes, which we take as evidence that behavioral clustering is a portable diagnosis primitive that generalizes beyond any single benchmark. We release the pipeline, three annotated corpora, and the cross-regime taxonomies.

Nicholas Saban

Patronus AI · University of California, Berkeley

LLM Evaluation Agent Evaluation Multi-Turn Dialogue Evaluation Language Model Probing

Jun 1, 2026cs.CL

Linear Probes Detect Task Format, Not Reasoning Mode in Language Model Hidden States

Linear probing of large language model (LLM) hidden states is widely used to claim that models learn distinct representations for different reasoning types. We test this by probing Qwen3-14B on three benchmarks spanning the classical trichotomy: LogiQA 2.0 (deductive), ARC-Challenge (inductive), and

α

NLI (abductive). At layer 32 of 40, linear probes achieve 100% cross-validated accuracy with well-separated geometry (intrinsic dimensionalities: 20.6, 28.5, 33.6; convex hull contamination

\leq

1.5%). However, this separation is entirely driven by format confounds. Residualizing source identity, option count, and response length reduces accuracy to chance. Trace-anchor similarity indicates largely shared reasoning across tasks (42.5% agreement vs.\ 33.3% chance), and causal steering with random controls (

n=20

) shows no functional link between geometry and reasoning mode (

p=0.286

). Thus, high probe accuracy reflects task format rather than computational structure, motivating routine format deconfounding in mechanistic interpretability.

Subramanyam Sahoo, Vinija Jain, Aman Chadha +1

Horizon Research · Meta · Apple +1

LLM Interpretability Representation Geometry in Language Models Language Model Probing Hidden-State Probing LLM Reasoning

May 30, 2026cs.CL

Detection vs. Execution: Single-Bucket Probes Miss Half the Mamba-2 State Sink

Mechanistic interpretability often assumes that probes identifying a representational signature also identify the circuit executing the corresponding computation. We show that this assumption can fail systematically in Mamba-2. Studying the state sink (disproportionate Delta-gate activation on boundary tokens, analogous to the attention sink), we find that single-bucket probes recover only a small execution layer while missing a much larger detection layer with the same representational signature. In Mamba-2, the state sink decomposes into two functional head sets. Single-bucket BOS-specialist heads (about 5% of heads at 2.7B) causally support both BOS-context and newline-target predictions across model scales and corpora. Dual heads (27-35% of heads, recovered by multi-class aggregation of the same probe) show stronger BOS-newline representational similarity but substantially weaker causal effects under ablation. Representational similarity does not imply functional equivalence. This distinction matters for downstream behaviour: ablating BOS-specialist heads collapses RULER NIAH retrieval accuracy from 1.00 to 0.00 at 1024 context length in both Mamba-1 2.8B and Mamba-2 2.7B, while size-matched complements preserve baseline performance. A random channel-bucketing control rules out substrate granularity alone, implicating Mamba-2's head-shared Delta projection. Probe-derived specialty can identify execution circuits; at coarse granularity the same probe also recovers detection circuits, and separating them requires class-conditional ablation rather than class-conditional cosine.

Yuhang Jiang

Independent Researcher

Mamba Attention Head Analysis Attention Sinks Language Model Probing Attention Head Specialization

May 30, 2026cs.LG

The Assistant as a Privileged Persona: A canonical reference in cross-persona self-recognition

Post-trained language models can recognize their own outputs from a sentence or two out of context. In a companion paper \citep{jack2026twomodes} we showed they can also recognize when they are currently acting on-policy, through the sharp entropy drop of assistant-mode generation. Both signals are tied to the Assistant persona that post-training mainly shapes. This paper widens the frame to cross-persona authorship judgement on Llama-3.1-70B-Instruct. We measure a matrix of authorship claim rates over a panel of evaluator and generator personas spanning librarian to dragon to Shakespeare, and make two claims. \emph{First}, on the Assistant's own row of the matrix, the Assistant's claim rate, the persona-vector distance from the Assistant in activation space, and the entropy gap between the Assistant's surprise on a persona's text and the persona's surprise on its own text are all tightly coupled. This extends the entropy signature of \emph{acting} from the companion paper to a retrospective signature of \emph{having acted}. \emph{Second}, this coupling fails off the Assistant's row: the natural symmetric extension of the entropy gap does not predict authorship for distinctive evaluators (pirate, dragon, Shakespeare); what does is asymmetric -- the evaluator's surprise compared to the Assistant's surprise on the same text, not to the generator's. We rule out the alternative that any persona could play this reference role by trying many candidate substitutes; none does. We interpret the asymmetry as the model performing an implicit Bayesian likelihood-ratio test against the Assistant as the canonical alternative hypothesis, with the persona-vector geometry of \citet{chen2025persona} (every persona a delta off the Assistant) ensuring that the Assistant is the only persona universally accessible to that test.

Asvin G

Institute for Advanced Study, Princeton

Personality Modeling in Language Models Authorship Attribution Persona Prompting Language Model Probing

May 30, 2026cs.LG

Hallucination Is Linearly Decodable from Mid-Layer Hidden States in Quantized LLMs

We investigate whether open-source LLMs encode a linearly separable truthfulness signal in their hidden states, and at which network depth this signal is strongest. Across three

7

B--

8

B instruction-tuned models (Llama-3.1-8B, Mistral-7B, Qwen2.5-7B) loaded in

4

-bit NF4 quantization, we extract per-layer hidden states on four hallucination benchmarks (TruthfulQA, HaluEval-QA, FEVER, and a controlled synthetic set) and compare four detection approaches: linear and MLP probes, INSIDE EigenScore, self-consistency, and attention entropy. A linear probe on a single mid-network layer achieves

0.904

--

1.000

AUROC on held-out splits, while sampling-based detectors do not exceed

0.541

AUROC under the same protocol. The truthfulness signal is approximately linear: MLP probes rarely surpass linear probes by more than

0.01

AUROC. Peak probing layers fall in a consistent band across model families on natural-language benchmarks -- blocks~

13

--

18

of~

32

for Llama and Mistral, and blocks~

19

--

25

of~

28

for Qwen. First-block attention entropy provides a complementary signal in knowledge-grounded settings (

0.866

--

0.941

AUROC on HaluEval-QA) at no additional inference cost. The low discriminability of sampling methods under this protocol reflects a structural mismatch between paired-label evaluation and the information these methods access, rather than an inherent limitation of those methods. Code and data are released for full reproducibility on a single

8

,GB GPU.

Aizierjiang Aiersilan

University of Macau

Hallucination in Language Models LLM Hallucination Detection Language Model Probing Hidden-State Probing

May 29, 2026cs.CL

What Am I Missing? Question-Answering as Hidden State Probing

Test-time reasoning has become a significant field of study since the introduction of chain-of-thought reasoning in large language models (LLMs). However, the mechanisms of this reasoning process are still under-explored -- from the same input prompt, and even the same partial solution, LLMs can produce varied answers if sampled multiple times. We propose to leverage question-asking as an inference-time intervention that articulates information about the model's hidden state. To achieve that, we present a student-teacher setting where a student asks questions to a teacher. We train a probe on the student's hidden state before and after asking a question and find it is predictive of the trajectory's final correctness, even before generating the teacher's answer. This suggests there is a meaningful signal from the self-diagnosis that occurs during question generation rather than information transfer from the teacher. We then frame question-asking as a sequential decision problem, using this probe as a quality score, and define a gating policy to ask questions that maximize likelihood of correctness. We find that the success of question-asking as an intervention is largely dependent on the model's self-consistency. Our empirical results show a gap between detection and recovery; while our gating policy captures model correctness and uncertainty, interventions are equally likely to harm correct trajectories as they are to recover incorrect ones. This gap between diagnosis and correction has broader implications on language models' capacity for self-refinement under uncertainty.

Chu Fei Luo, Samuel Dahan, Xiaodan Zhu

Department of Electrical and Computer Engineering & Ingenuity Labs, Queen’s University · Conflict Analytics Lab, Queen’s University · Cornell Law School +1

LLM Self-Refinement Language Model Self-Assessment LLM Uncertainty Estimation Metacognition in Language Models Language Model Probing LLM Reasoning

May 29, 2026cs.SE

Knowledge boundary probing and demand-guided intervention for LLM-based power system code generation

Large language models (LLMs) can turn grid-analysis requests into executable programs for power-system simulation, but utilities and research laboratories often require on-premise deployment. In this setting, first-pass failures frequently arise at an API-knowledge boundary, through hallucinated functions, misused parameters, and mishandled result tables. We present PowerCodeBench, a parameterised benchmark generator released as a frozen 2,000-task suite for pandapower, and a deployment-time workflow that requires no weight updates. Documentation-driven L0-L3 probes produce per-model API profiles for diagnosis, model comparison, documentation allocation, and backend calibration. A query-side demand estimator selects layered API evidence before generation, while execution feedback routes targeted repair. Across ten open-weight LLMs (1.5B-480B) and four mid-tier APIs, the validation-enabled workflow raises scalar-match accuracy by 32-56 percentage points after up to three repair rounds relative to an unassisted first pass, for every model of at least 7B and every API. Open-weight models in the 70B-120B range reach the four-vendor mid-tier accuracy range under matched no-tool conditions. Selective injection approaches the full-layer reference using 41% of its prompt tokens. Among model-item pairs passing numerical checks under both workflows, engineering review confirms the requested analysis in 88% of full-workflow outputs versus 66% under plain repair. Round-0 pilots on OpenDSS and PyPSA motivate staged onboarding from broad retrieval at cold start to calibrated selective injection. Measured throughput, latency, energy, and allocated GPU memory establish a practical on-premise serving envelope.

Hui Wu, Xiaoyang Wang, Zhong Fan

Department of Engineering, University of Exeter · Department of Computer Science, University of Exeter

Automated Program Repair Code Generation Code Generation Benchmarks LLM Code Generation Language Model Probing

May 28, 2026cs.CL

Probing the Prompt KV Cache: Where It Becomes Dispensable

Prior KV cache compression schemes empirically demonstrate that the prompt cache is partially redundant during decoding, dropping or summarising entries with little accuracy loss. We ask when and what kind of redundancy: at which layers, after how many decoding steps, and in what form can the prompt span KV cache be replaced without breaking the task. A controlled splice intervention swept over layer cutoff and decoding steps shows this redundancy is about form (chat template scaffolding) rather than content. Replacing the upper layer prompt span KV cache with KV cache from a chat template scaffold whose user content is a neutral filler recovers near clean accuracy, while zeroing the same slots collapses accuracy. The dissociation replicates across the Qwen3, Gemma 3, and Llama 3 families on multiple datasets.

Vinayshekhar Bannihatti Kumar, Manoj Ghuhan Arivazhagan, Disha Makhija +1

AWS AI Labs

KV Caching KV-Cache Management KV-Cache Compression Causal Interventions in Language Models Language Model Probing

May 28, 2026cs.CL

Latent Performance Profiling of Large Language Models

Large language models (LLMs) frequently achieve impressive scores on standardized benchmarks, yet accuracy alone offers a limited view of their capabilities. Evaluating open-source LLMs on leaderboards faces persistent issues such as data contamination, a narrow task scope, and poor alignment with real-world reliability. Benchmark-based evaluations such as MMLU-Pro, BBH, or IFEval primarily capture \textit{what} a model outputs on fixed test sets, not \textit{how} it processes information, calibrates uncertainty, or structures internal knowledge. In this article, we advocate for a shift from benchmark-centric evaluation toward a complementary, \textit{state-centered intrinsic assessment} of LLMs. To this end, we introduce \textbf{Latent Performance Profiling (LPP)} --- a framework that derives task-agnostic diagnostics from hidden activations and output distributions. LPP defines a set of scalar metrics on a model's latent representations and dynamics, revealing traits that enable interpretable comparisons and uncover hidden vulnerabilities. Unlike static accuracy scores, LPP provides stable, architecture-sensitive signatures across models of similar size. With extensive empirical analyses across eight LLMs, spanning a size range of 0.5B-14B, we demonstrate that models with similar benchmark scores can exhibit contrasting latent profiles, such as differences in entropy or compressibility. Guided by these insights, we design synthetic probes for uncertainty and symbolic reasoning that align with intrinsic metrics while decoupling from leaderboard bias. We recommend reporting LPP alongside benchmarks to provide a deeper, interpretable understanding of model behavior, enabling more reliable model selection, safety assessment, and evaluation beyond surface-level accuracy.

Tanmoy Chakraborty, Ayan Sengupta, Suparna Bhattacharya +7

Department of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi India · Yardi School of Artiﬁcial Intelligence, Indian Institute of Technology Delhi, New Delhi India · Hewlett Packard Enterprise, India +5

LLM Evaluation LLM Interpretability Language Model Probing

May 27, 2026cs.CL

Error as a Lens: Probing LLM Reasoning through Synthetic Misconception Generation

Personalized tutoring, teacher training, and education research need access to \emph{targeted} synthetic misconceptions, but privacy and IRB constraints make labelled corpora of real student errors scarce. LLMs could in principle generate synthetic errors at scale, but producing an arbitrary wrong answer is easy for a modern LLM while producing one that matches a specified cognitive failure mode is much harder. We present a framework that generates errors targeted to a five-class taxonomy adapted from the revised Bloom's taxonomy, evaluated on questions from the TheoremQA dataset. A Generation Agent (GA) drafts a candidate erroneous solution conditioned on a target class, and an Examination Agent (EA) judges whether the draft is incorrect and class-consistent. The framework yields a reusable recipe for building class-stratified synthetic error datasets where authentic student corpora are unavailable. As a secondary diagnostic, targeted error generation is substantially harder than free-form incorrect-answer generation, and answer-grounding contributes more than expanded examples or external textbook content.

Xinming Yang, Jun Li

CUNY Graduate Center · CUNY Queens College

Synthetic Data Generation Generative AI in Education Language Model Probing

May 27, 2026cs.CL

Reverse Probing: Supervised Token-level Uncertainty Quantification for Large Language Models in Clinical Text

As large language models are increasingly deployed for clinical text, ensuring they can reliably signal their own uncertainty becomes critical. Most existing uncertainty quantification (UQ) methods are designed for open-domain generation and cannot localize uncertainty at the token or span level in long clinical text. We propose Reverse Probing, the first UQ framework specialized for clinical summarization, which estimates token-level uncertainty directly from pre-existing labeled summaries. Rather than sampling new outputs, Reverse Probing treats the text as a probe into the model's internal state, extracting uncertainty signals from four categories of internal activations. We evaluate on two expert-annotated clinical datasets and outperform eight adapted baselines on all metrics, achieving up to 4 times higher AUPRC while reducing inference time and computational costs. Feature analysis reveals that delta energy and neighborhood context are the most consistent predictors across all models. This study offers interpretable insights into how models internally respond to unsupported clinical content.

Bushi Xiao, Sarvesh Soni, Daisy Zhe Wang

University of Florida Gainesville, FL · U.S. National Library of Medicine Bethesda, MD

Text Summarization LLM Uncertainty Estimation Medical LLMs Language Model Probing