arXiv:2602. 04718v4 Announce Type: replace Abstract: A central premise in mechanistic interpretability is that meaningful concepts in language models are represented by linear features in activation space.
arXiv:2602. 04718v2 Announce Type: replace-cross Abstract: A central premise in mechanistic interpretability is that meaningful concepts in language models are represented by linear features in activation space.
arXiv:2608. 10537v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) have helped uncover mechanistic explanations for LLM behaviours such as reasoning, jailbreaking etc.
arXiv:2601.08058v2 Announce Type: replace-cross Abstract: Chain-of-Thought (CoT) prompting often improves the reasoning performance of large language models (LLMs), but the internal signal that trigg...
The paper proposes that self‑improvement methods for language models—such as debate, bootstrap, and internal coherence maximization—can be unified under a framework called coherence optimization. This framework seeks a context‑to‑behavior mapping that is most compressible and jointly predictable, and the authors prove its equivalence to description‑length regularization. They further show that coherence regularization, when guided by a pretrained model prior, optimizes a lower bound on worst‑case accuracy for semi‑supervised learning, offering a theoretical explanation for the success of feedback‑free self‑improvement and predictions about its limits.
arXiv:2606. 24964v1 Announce Type: new Abstract: Understanding the features of large language models (LLMs) is a central goal of interpretability.
arXiv:2508.20443v3 Announce Type: replace Abstract: Large language models (LLMs) are trained on massive datasets that may include private or copyrighted content. Due to growing privacy and ownership...
arXiv:2606. 19946v1 Announce Type: cross Abstract: Activation steering controls model behavior by modifying intermediate hidden states at inference time without retraining.
The wide-scale use of sparse autoencoders (SAEs) as interpretability tools is limited by inconsistent links between SAE features and model behavior. Features with clear activation descriptions may have weak or unexpected causal effects; steering can vary across prompts or oppose the intended direction; and activation-based feature selection can miss features that produce the desired output change.
arXiv:2511. 09432v2 Announce Type: replace Abstract: Machine learning (ML) models achieve remarkable performance but remain hard to interpret due to their scale and complexity.
arXiv:2608.30462v1 Announce Type: cross Abstract: Large language models exhibit substantial performance variation across languages, even when solving semantically equivalent tasks. Existing analyses...
arXiv:2607. 24645v1 Announce Type: cross Abstract: The wide-scale use of sparse autoencoders (SAEs) as interpretability tools is limited by inconsistent links between SAE features and model behavior.
Sparse autoencoders (SAEs) are proposed to extract numerous features from large language model (LLM) representations, yet explaining these features still relies primarily on external observation. This reliance leads to superficial explanations inferred from observed model behavior and computational inefficiency from collecting such behavioral evidence at scale.