Mechanisms of Introspective Awareness
This paper studies how models detect steering vectors and similar internal interventions. It asks whether the behavior reflects genuine introspective circuitry or simpler heuristics.
From: Proceedings of the 43rd International Conference on Machine Learning
The authors analyze open-source models to understand mechanisms behind introspection-like behavior.
They find the ability is robust and tied to distributed internal computation rather than a single obvious feature.
The paper does not show subjective awareness, and its results stay within mechanistic detection tasks.
Source: PMLR ↗. This is a summary; the source is the authority.
This strengthens mechanistic evidence for self-monitoring-like behavior, but it still falls far short of evidence for conscious experience.
How we weigh evidence like this: the evidence levels.
- Mechanistic Interpretability Should Prioritize Feature Consistency in Sparse AutoencodersThis paper argues that sparse autoencoders should be judged not only by reconstruction quality and sparsity but also by whether the same fea…
- Mechanistic Interpretability for Neural NetworksThis is a broad survey of mechanistic interpretability, covering the goal of reverse-engineering internal algorithms in neural networks. It …
- Constructing Interpretable Features from Compositional Neuron Representations in Language ModelsThis paper proposes an alternative to standard sparse autoencoders by decomposing MLP activations with semi-nonnegative matrix factorization…
- Interpretability Transfer from Language to Vision via Sparse Autoencoder Transfer AlignmentThis work transfers labeled sparse-autoencoder features from language to vision inside a multimodal model. It uses those features to make vi…