Beyond Behavioural Trade-Offs: Mechanistic Tracing of Pain-Pleasure Decisions in an LLM

• URL: http://arxiv.org/abs/2602.19159v1
• Date: Sun, 22 Feb 2026 12:42:38 GMT
• Title: Beyond Behavioural Trade-Offs: Mechanistic Tracing of Pain-Pleasure Decisions in an LLM
• Authors: Francesca Bianco, Derek Shiller,
• Abstract summary: Prior work suggests that some LLMs alter choices when options are framed as causing pain or pleasure, and that such deviations can scale with stated intensity.<n>We investigate how valence-related information is represented and where it is causally used inside a transformer.<n>Using Gemma-2-9B-it and a minimalist decision task modelled on prior work, we (i) map representational availability with layer-wise linear probing across streams.<n>We (ii) test causal contribution with activation interventions (steering; patching/ablation), and (iii) quantify dose-response effects over an epsilon grid.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Prior behavioural work suggests that some LLMs alter choices when options are framed as causing pain or pleasure, and that such deviations can scale with stated intensity. To bridge behavioural evidence (what the model does) with mechanistic interpretability (what computations support it), we investigate how valence-related information is represented and where it is causally used inside a transformer. Using Gemma-2-9B-it and a minimalist decision task modelled on prior work, we (i) map representational availability with layer-wise linear probing across streams, (ii) test causal contribution with activation interventions (steering; patching/ablation), and (iii) quantify dose-response effects over an epsilon grid, reading out both the 2-3 logit margin and digit-pair-normalised choice probabilities. We find that (a) valence sign (pain vs. pleasure) is perfectly linearly separable across stream families from very early layers (L0-L1), while a lexical baseline retains substantial signal; (b) graded intensity is strongly decodable, with peaks in mid-to-late layers and especially in attention/MLP outputs, and decision alignment is highest slightly before the final token; (c) additive steering along a data-derived valence direction causally modulates the 2-3 margin at late sites, with the largest effects observed in late-layer attention outputs (attn_out L14); and (d) head-level patching/ablation suggests that these effects are distributed across multiple heads rather than concentrated in a single unit. Together, these results link behavioural sensitivity to identifiable internal representations and intervention-sensitive sites, providing concrete mechanistic targets for more stringent counterfactual tests and broader replication. This work supports a more evidence-driven (a) debate on AI sentience and welfare, and (b) governance when setting policy, auditing standards, and safety safeguards.

Related papers

• Hallucination Begins Where Saliency Drops [18.189047289404325]
  hallucinations frequently arise when preceding output tokens exhibit low saliency toward the prediction of the next token.<n>We introduce LVLMs-Saliency, a gradient-aware diagnostic framework that quantifies the visual grounding strength of each output token.<n>Our method significantly reduces hallucination rates while preserving fluency and task performance, offering a robust and interpretable solution.
  arXiv  Detail & Related papers  (2026-01-28T05:50:52Z)
• Ground What You See: Hallucination-Resistant MLLMs via Caption Feedback, Diversity-Aware Sampling, and Conflict Regularization [38.469173375694076]
  This paper systematically analyzes the root causes of hallucinations in Multimodal Large Language Models (MLLMs)<n>It identifies three critical factors: (1) an over-reliance on chained visual reasoning, where inaccurate initial descriptions anchor subsequent inferences to incorrect premises; (2) insufficient exploration diversity during policy optimization, leading the model to generate overly confident but erroneous outputs; and (3) destructive conflicts between training samples, where NTK similarity causes false associations and unstable parameter updates.<n> Experimental results demonstrate that our proposed method significantly reduces hallucination rates and effectively enhances the inference accuracy of MLLMs.
  arXiv  Detail & Related papers  (2026-01-09T07:59:18Z)
• Anchors in the Machine: Behavioral and Attributional Evidence of Anchoring Bias in LLMs [0.0]
  This paper advances the study of anchoring in large language models (LLMs) through three contributions.<n>Results reveal robust anchoring effects in Gemma-2B, Phi-2, and Llama-2-7B, with attribution signaling that the anchors influence reweighting.<n>Findings demonstrate that anchoring bias in LLMs is robust, measurable, and interpretable, while highlighting risks in applied domains.
  arXiv  Detail & Related papers  (2025-11-07T23:35:19Z)
• Where Did It Go Wrong? Attributing Undesirable LLM Behaviors via Representation Gradient Tracing [12.835224376066769]
  Large Language Models (LLMs) have demonstrated remarkable capabilities, yet their deployment is frequently undermined by undesirable behaviors.<n>We introduce a novel and efficient framework that diagnoses a range of undesirable LLM behaviors by analyzing representation and its gradients.<n>We systematically evaluate our method for tasks that include tracking harmful content, detecting backdoor poisoning, and identifying knowledge contamination.
  arXiv  Detail & Related papers  (2025-09-26T12:07:47Z)
• GrAInS: Gradient-based Attribution for Inference-Time Steering of LLMs and VLMs [56.93583799109029]
  GrAInS is an inference-time steering approach that operates across both language-only and vision-language models and tasks.<n>During inference, GrAInS hidden activations at transformer layers guided by token-level attribution signals, and normalizes activations to preserve representational scale.<n>It consistently outperforms both fine-tuning and existing steering baselines.
  arXiv  Detail & Related papers  (2025-07-24T02:34:13Z)
• Zero-Shot Event Causality Identification via Multi-source Evidence Fuzzy Aggregation with Large Language Models [11.541829239773643]
  Event Causality Identification (ECI) aims to detect causal relationships between events in textual contexts.<n>Existing ECI models predominantly rely on supervised methodologies, suffering from dependence on large-scale annotated data.<n>We propose MEFA, a novel zero-shot framework based on Multi-source Evidence Fuzzy Aggregation.
  arXiv  Detail & Related papers  (2025-06-06T01:56:05Z)
• 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)
• SEAL: Steerable Reasoning Calibration of Large Language Models for Free [58.931194824519935]
  Large Language Models (LLMs) have demonstrated compelling capabilities for complex reasoning tasks via the extended chain-of-thought (CoT) reasoning mechanism.<n>Recent studies reveal substantial redundancy in the CoT reasoning traces, which negatively impacts model performance.<n>We introduce SEAL, a training-free approach that seamlessly calibrates the CoT process, improving accuracy while demonstrating significant efficiency gains.
  arXiv  Detail & Related papers  (2025-04-07T02:42:07Z)
• Unifying Perplexing Behaviors in Modified BP Attributions through Alignment Perspective [61.5509267439999]
  We present a unified theoretical framework for methods like GBP, RectGrad, LRP, and DTD.<n>We demonstrate that they achieve input alignment by combining the weights of activated neurons.<n>This alignment improves the visualization quality and reduces sensitivity to weight randomization.
  arXiv  Detail & Related papers  (2025-03-14T07:58:26Z)
• Understanding, Predicting and Better Resolving Q-Value Divergence in Offline-RL [86.0987896274354]
  We first identify a fundamental pattern, self-excitation, as the primary cause of Q-value estimation divergence in offline RL. We then propose a novel Self-Excite Eigenvalue Measure (SEEM) metric to measure the evolving property of Q-network at training. For the first time, our theory can reliably decide whether the training will diverge at an early stage.
  arXiv  Detail & Related papers  (2023-10-06T17:57:44Z)
• Identifying Weight-Variant Latent Causal Models [82.14087963690561]
  We find that transitivity acts as a key role in impeding the identifiability of latent causal representations. Under some mild assumptions, we can show that the latent causal representations can be identified up to trivial permutation and scaling. We propose a novel method, termed Structural caUsAl Variational autoEncoder, which directly learns latent causal representations and causal relationships among them.
  arXiv  Detail & Related papers  (2022-08-30T11:12:59Z)

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.