Energy-navigated trajectory shaping during training produces 8-step discrete flow matching students that achieve 32% lower perplexity than 1024-step teachers on 170M language models with unchanged inference cost.
The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale , url =
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3roles
background 1polarities
unclear 1representative citing papers
GUARD-IT performs machine unlearning in LLMs via input-dependent activation steering at inference time, matching or exceeding gradient-based baselines on TOFU and MUSE while preserving utility and working under quantization.
LLMs exhibit compartmentalization by learning separate internal representations for equivalent concepts presented differently, which reduces sample efficiency and resists unification even with synthetic parallel data.
citing papers explorer
-
Trajectory as the Teacher: Few-Step Discrete Flow Matching via Energy-Navigated Distillation
Energy-navigated trajectory shaping during training produces 8-step discrete flow matching students that achieve 32% lower perplexity than 1024-step teachers on 170M language models with unchanged inference cost.
-
Inference-Time Machine Unlearning via Gated Activation Redirection
GUARD-IT performs machine unlearning in LLMs via input-dependent activation steering at inference time, matching or exceeding gradient-based baselines on TOFU and MUSE while preserving utility and working under quantization.
-
Language models struggle with compartmentalization
LLMs exhibit compartmentalization by learning separate internal representations for equivalent concepts presented differently, which reduces sample efficiency and resists unification even with synthetic parallel data.