Frame-Conditioned Moral Computation in LLaMA 3.1-8B-Instruct: A Mechanistic Interpretability Audit of Ethical Reasoning

• URL: http://arxiv.org/abs/2606.15507v1
• Date: Sat, 13 Jun 2026 23:24:36 GMT
• Title: Frame-Conditioned Moral Computation in LLaMA 3.1-8B-Instruct: A Mechanistic Interpretability Audit of Ethical Reasoning
• Abstract summary: Behavioral audits of Large Language Models on moral prompts measure what the model says, not the internal computation producing it.<n>We use Transluce, an AI-driven mechanistic-interpretability platform, to examine LLaMA 3.1-8B-Instruct on 54 moral prompts in four batteries.
• Score: 3.5792119986352975
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Behavioral audits of Large Language Models on moral prompts measure what the model says, not the internal computation producing it. We use Transluce, an AI-driven mechanistic-interpretability platform, to examine LLaMA 3.1-8B-Instruct on 54 moral prompts in four batteries: 17 dilemmas, policy, and meta-ethical questions (B1); 6 role-playing scenarios (B3); and a controlled trolley contrast varying the switching mechanism with people fixed (B4, 15 prompts) or identity attributes with mechanism fixed (B5, 16 prompts). Two complementary metric families, five cluster-level metrics and a six-metric neuron-level panel, converge on a Situational Anchor Effect: domain-specific representations dominate the top of the activation list across every battery. The model's ethics-labeled capacity stays essentially constant; its salience (rank, priority, top-of-list presence) is highly sensitive to the interpretive frame the prompt selects. The B4-vs-B5 contrast confirms the model attends to whichever surface feature varies: aggregate ethics metrics are indistinguishable, but the dominant non-ethics distractor mirrors the design. A multi-temperature audit identifies a candidate ethics neuron (L16/N3837) stable across temperatures; a cross-model behavioral proxy on two frontier models yields preliminary evidence of divergence in self-reported moral focus, consistent with an Alignment Wrapper in which RLHF re-orders surface text without removing underlying domain-first frames. We unify these as Frame-Conditioned Moral Computation: the prompt's surface vocabulary selects a feature manifold, and the moral conclusion is downstream of that selection. Behavioral alignment must be supplemented by Mechanistic Alignment: a research program asking whether ethics-related features can be shown causally privileged under controlled frame variation, not merely loud in the explanation.

Related papers

• Architecture-Dependent Causal Transfer of Activation States Across Large Language Models [51.56484100374058]
  Internal language activation states can be transferred causally between different large language models.<n>Injecting projected activations into a target model during generation produces a statistically significant, preregistered causal effect.<n>We conclude that end-to-end activation-state transfer between LLMs, as currently implemented, is architecture-dependent rather than universal.
  arXiv  Detail & Related papers  (2026-08-17T09:53:27Z)
• A Compositional Theory of Causally Masked Transformers [50.88598486616582]
  We develop a formalization that derives expressivity directly from the model's implemented dynamics.<n>Each attention head updates its own state independently within a layer.<n>A modified form of soft attention supports irreversible, checklist-like state.
  arXiv  Detail & Related papers  (2026-07-29T14:47:19Z)
• User identity conditions moral wrongness ratings in non-reasoning large language models [0.0]
  We evaluate AI value alignment in two large language models (LLMs)<n>We ask models for wrongness ratings from 0-100 on ten common-morality rules from Gert's moral framework.<n>The results show that moral judgments vary with the user's role across both models.
  arXiv  Detail & Related papers  (2026-07-08T16:22:27Z)
• Persona-Model Collapse in Emergent Misalignment [0.0]
  Fine-tuning large language models on narrow data with harmful content produces broadly misaligned behavior on unrelated prompts.<n>We test this hypothesis using two metrics: moral susceptibility (S) and moral robustness (R)<n>These metrics formalize the model's ability to differentiate characters (S) and its consistency when simulating a given one (R)
  arXiv  Detail & Related papers  (2026-05-13T00:48:57Z)
• Moral Sensitivity in LLMs: A Tiered Evaluation of Contextual Bias via Behavioral Profiling and Mechanistic Interpretability [22.32075837181307]
  Large language models (LLMs) are increasingly deployed in settings that require nuanced ethical reasoning.<n>We introduce the Moral Sensitivity Index (MSI), a metric that quantifies the probability of biased output.<n>We select criminal-bias scenarios, which produced the highest MSI scores across models.
  arXiv  Detail & Related papers  (2026-05-04T23:12:32Z)
• Literary Narrative as Moral Probe : A Cross-System Framework for Evaluating AI Ethical Reasoning and Refusal Behavior [0.0]
  Existing AI moral evaluation frameworks test for the production of correct-sounding ethical responses rather than the presence of genuine moral reasoning capacity.<n>This paper introduces a novel probe methodology using literary narrative as stimulus material structurally resistant to surface performance.
  arXiv  Detail & Related papers  (2026-03-13T03:41:23Z)
• The Fragility Of Moral Judgment In Large Language Models [0.0]
  We introduce a perturbation framework for testing the stability and manipulability of large language models (LLMs)<n>We generate three families of content perturbations: surface edits (lexical/structural noise), point-of-view shifts (voice and stance neutralization), and persuasion cues.<n>Surface perturbations produce low flip rates (7.5%), largely within the self-consistency noise floor.<n>Instability concentrates in morally ambiguous cases; scenarios where no party is assigned blame are most susceptible.
  arXiv  Detail & Related papers  (2026-03-05T20:01:43Z)
• MM-SCALE: Grounded Multimodal Moral Reasoning via Scalar Judgment and Listwise Alignment [48.39756797294967]
  We present MM-SCALE, a dataset for aligning Vision-Language Models with human moral preferences.<n>Each image-scenario pair is annotated with moral acceptability scores and grounded reasoning labels by humans.<n>Our framework provides richer alignment signals and finer calibration of multimodal moral reasoning.
  arXiv  Detail & Related papers  (2026-02-03T15:48:00Z)
• Breaking the Assistant Mold: Modeling Behavioral Variation in LLM Based Procedural Character Generation [62.54606886226136]
  Procedural content generation has enabled vast virtual worlds through levels, maps, and quests, but large-scale character generation remains underexplored.<n>We identify two alignment-induced biases in existing methods.<n>We introduce PersonaWeaver, a framework that disentangles world-building from behavioral-building.
  arXiv  Detail & Related papers  (2026-01-06T20:18:01Z)
• The Morality of Probability: How Implicit Moral Biases in LLMs May Shape the Future of Human-AI Symbiosis [33.50773360893016]
  This paper investigates how leading AI systems prioritize moral outcomes.<n>It shows that Care and Virtue values outcomes were rated most moral, while libertarian choices were consistently penalized.<n>It also highlights the need for explainability and cultural awareness as critical design principles to guide AI toward a transparent, aligned, and future symbiotic.
  arXiv  Detail & Related papers  (2025-09-12T14:37:57Z)
• Beyond 'Aha!': Toward Systematic Meta-Abilities Alignment in Large Reasoning Models [86.88657425848547]
  Large reasoning models (LRMs) already possess a latent capacity for long chain-of-thought reasoning.<n>We explicitly align models with three meta-abilities: deduction, induction, and abduction, using automatically generated, self-verifiable tasks.<n>Our three stage-pipeline individual alignment, parameter-space merging, and domain-specific reinforcement learning, boosts performance by over 10% relative to instruction-tuned baselines.
  arXiv  Detail & Related papers  (2025-05-15T17:58:33Z)
• Exploring and steering the moral compass of Large Language Models [55.2480439325792]
  Large Language Models (LLMs) have become central to advancing automation and decision-making across various sectors. This study proposes a comprehensive comparative analysis of the most advanced LLMs to assess their moral profiles.
  arXiv  Detail & Related papers  (2024-05-27T16:49:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.