Multiple semantically adjacent truth directions make models more vulnerable to manipulations that shift outputs without obvious signs in the primary truth direction

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• We hypothesize that explicitly instructing the model to evaluate the correctness of the given statement may change the geometry of truth directions.hypothesis0.833
  Motivating hypothesis for Section 5's investigation of prompt template effects.
• Universality claims for truth directions are more limited than previously assumed, with significant differences observable for various model layers, task difficulties, task types, and prompt templates.claim0.802
  Overarching conclusion summarizing the paper's contribution relative to prior universality claims.
• Truth directions fail to generalize to harder tasks (F3-F5) regardless of prompt template because activations remain highly entangled.claim0.799
  Establishes task difficulty as a hard limit that instructions cannot overcome.
• Does instructing the model to assess correctness affect the geometry of truth directions?question0.792
  One of the three guiding research questions of the paper.
• No single layer is universally optimal for probing truth directions; different tasks peak at different layers.claim0.792
  Argues against the single-layer analysis approach of prior work.
• Are the discovered truth directions robust to architectural variation and fine-tuning differences across model families?question0.791
  Open question on generalization beyond Gemma and Qwen families
• The need for genuine counting over lists of more than two elements introduces the key limitation of truth directions.claim0.790
  Identified as the exact computational operation that breaks truth direction generalization.
• Truth directions emerge in earlier layers for factual tasks and later layers for arithmetic tasks.claim0.788
  Core empirical claim about the layer-dependence of truth direction emergence as a function of task type.