Sympatheia is a speech-to-speech dialogue framework that generates emotionally adaptive responses by conditioning its output on continuous affect signals derived from user speech or external multimoda…

The paper theoretically analyzes the properties that optimal sparse autoencoder (SAE) dictionaries must satisfy, deriving constraints that explain observed SAE behaviors like hierarchical splitting an…

This paper demonstrates that Sparse Autoencoders (SAEs) can effectively steer Large Language Models (LLMs) on the AxBench benchmark, achieving performance comparable to LoRA baselines when combined wi…

The paper demonstrates that the valence structure learned by modern LLMs aligns with human EEG emotional representations, but finds that further supervised alignment is ineffective due to a phenomenon…

The paper demonstrates that sparse autoencoders can successfully extract a large set of interpretable, causally influential features from the production-scale Claude 3 Sonnet language model.

The paper demonstrates that integrating Sparse Autoencoders (SAEs) into transformer residual streams significantly enhances the robustness of Large Language Models against various jailbreak attacks by…

The paper introduces Residualized Sparse Autoencoders (ReSAEs) to improve multi-layer interventions in transformers by training each layer on the residual activation, which better preserves cross-laye…

The paper introduces residualized temporal Sparse Autoencoders (SAEs) to analyze the full spatiotemporal structure of activations generated during the iterative denoising process of diffusion models,…

The paper introduces a distributional framework using Wasserstein distance to unify the semantic comparison of sparse autoencoder features across different layers and to automatically compress large f…

MindVoice is a neuro-to-speech framework that uses pretrained priors to disentangle and reconstruct intelligible speech from noisy, non-invasive neural signals, significantly outperforming existing me…

The paper proposes an inference-time activation steering framework, utilizing orthogonalization, to achieve fine-grained, deterministic control over discrete musical attributes like Pitch and Duration…