Goto

Collaborating Authors

 Franklin County


Unsupervised decoding of encoded reasoning using language model interpretability

Fang, Ching, Marks, Samuel

arXiv.org Artificial Intelligence

As large language models become increasingly capable, there is growing concern that they may develop reasoning processes that are encoded or hidden from human oversight. To investigate whether current interpretability techniques can penetrate such encoded reasoning, we construct a controlled testbed by fine-tuning a reasoning model (DeepSeek-R1-Distill-Llama-70B) to perform chain-of-thought reasoning in ROT-13 encryption while maintaining intelligible English outputs. We evaluate mechanistic interpretability methods--in particular, logit lens analysis--on their ability to decode the model's hidden reasoning process using only internal activations. We show that logit lens can effectively translate encoded reasoning, with accuracy peaking in intermediate-to-late layers. Finally, we develop a fully unsupervised decoding pipeline that combines logit lens with automated paraphrasing, achieving substantial accuracy in reconstructing complete reasoning transcripts from internal model representations. These findings suggest that current mechanistic interpretability techniques may be more robust to simple forms of encoded reasoning than previously understood. Our work provides an initial framework for evaluating interpretability methods against models that reason in non-human-readable formats, contributing to the broader challenge of maintaining oversight over increasingly capable AI systems.


The Next Decade in AI: Four Steps Towards Robust Artificial Intelligence

Marcus, Gary

arXiv.org Artificial Intelligence

Recent research in artificial intelligence and machine learning has largely emphasized general-purpose learning and ever-larger training sets and more and more compute. In contrast, I propose a hybrid, knowledge-driven, reasoning-based approach, centered around cognitive models, that could provide the substrate for a richer, more robust AI than is currently possible.


K & K Technical (@KK_Technical)

#artificialintelligence

Are you sure you want to view these Tweets? How will self-driving cars disrupt the auto industry? Get the inside track and land your new job with these career seeking strategies http://goo.gl/rO8Qch We hope the tragedy in Tianjin reminds all of us in manufacturing, to carefully adhere to all safety procedures http://goo.gl/MEbQr6 More and more engineers are emerging as successful business leaders across the U.S. http://goo.gl/wT8smy