EMNLP 2026 paper analyzes parametric-context knowledge interaction via Rank-2 subspace disentanglement
IAugenstein · x · 2026-09-02
A paper from Isabelle Augenstein's group, "Multi-Step Knowledge Interaction Analysis via Rank-2 Subspace Disentanglement," was accepted to EMNLP 2026 Findings. It studies the interaction dynamics between parametric knowledge (PK) and contextual knowledge (CK) when models generate longer natural language explanations, using a Rank-2 subspace disentanglement approach.
More from Research
- Humanity's First Interstellar Spacecraft to Alpha Centauri Sparks Buzz — rand_longevity · 2026-09-02
- KV Cache as an Information Bottleneck Between Tokens: Austin Huang's Idea — austinvhuang · 2026-09-02
- Nature Comms Perspective links explainable AI, foundation models and scientific discovery — ricardovinuesa · 2026-09-02
- Looped transformer is no dark art: rasbt debunks the OpenAI Astra rumor — rasbt · 2026-09-02
- A decade of robot skill learning: from mass imitation to few-shot RL — stepjamUK · 2026-09-02
- GAM repurposes a geometric foundation model as robot manipulation policy, moving beyond 2D VLA — rsasaki0109 · 2026-09-02