Same Outcome, Different Readout: What Does a Steerable Valence Direction in LLMs Represent?
cs.LG, cs.AI
Submitted: 2026-09-19
Updated: 2026-09-27
Comments: 18 pages, 13 figures, 10 tables
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Gemma 3 Technical Report
- How's it going? Reinforcement learning in language models recruits a functional welfare axis
- Functional Emotions or Situational Contexts? A Discriminating Test from the Mythos Preview System Card
- Qwen3 Technical Report
- Emotion Concepts and their Function in a Large Language Model
- The Pain Axis: LLMs Represent Self-Directed Harm and Act on It
- Sparse Reward Subsystem in Large Language Models
- Representation Engineering: A Top-Down Approach to AI Transparency
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks