cs.CLNov 14, 2026

A Bag of Concepts? Testing Role-Filler Binding in the Verbalizable Workspace

Authors: Reya Dawlah, Elizabeth McMillin, Michael Shihong Zhang, Badrish Ananth, Rishi Adi Ruia, Ane Zuñiga, Thomas Jiralerspong

Organizations: University of Texas at Dallas · MiraCosta College · University of Waterloo · University of California, Irvine · 

Abstract

Runtime alignment monitors built on a verbalizable workspace have been proposed as a way to surface a model's silent cognition before it produces output. Their value rests on the workspace carrying enough structure to distinguish safety-relevant propositions, most of which are relational: "the model plans to deceive the user" and "the user plans to deceive the model" share a concept set and differ only in role assignment. We ask whether the workspace carries the role-filler bindings (which entity fills which role) such a monitor would need, on three frontier language models (Qwen3.6-27B, Gemma-3-12B, Gemma-3-27B-IT), under two lens families (J-lens and R-lens), and with two methods for fitting the role direction (difference-of-means and LRE-gradient). We find a clean dissociation: the workspace carries directions that steer what the model outputs about roles but not the representations the model uses to compute them. Linear probes recover role at near-ceiling from the residual stream but score below a rank-matched random control when reading the workspace, and steering the model along the fitted role direction changes which entity it binds to no more than strength-matched control edits do. Ablating the workspace (zeroing the residual stream's projection onto its ten most active lens directions) damages binding on one of the three models (Gemma-3-27B-IT, holding across three lens variants) and shows no binding-specific damage on the other two, with a rank-matched random subspace as the control; this confirms our design detects binding when present, though scale and instruction tuning are confounded there. A workspace-lens monitor built on these directions would therefore miss a role assignment the model itself represents.

Figures & tables

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. What Roles Do System Prompts Really Assign? Validating and Steering Occupational Role Vectors in LLMs

    Nov 11, 2026Jose L. Sampedro Mazon, Ilmi Yoon, Nadim SaadActivation SteeringLLM Interpretability

  2. In-Context Binding Capacity in Language Models

    Sep 24, 2026Manas Venkata Sai Ravulapalli, Samrath Singh ChadhaWorking Memory in Language ModelsLanguage Model Scaling Laws

  3. Not All LLM Reasoning is Visible in the Chain-of-Thought

    Jul 24, 2026Vatsal Baherwani, Tom Goldstein, Ashwinee PandaCoT FaithfulnessLLM Interpretability