FIELD NOTES TOPIC
computational efficiency
3 notes, newest first.
Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL Loss
The researchers present a study to make knowledge distillation training efficient for small language models deployed under tight constraints. They propose two methods: offline knowledge distillation (caching teacher's…
Fast Multi-dimensional Refusal Subspaces via RFM-AGOP
A new method called Fast Multi-dimensional Refusal Subspaces via RFM-AGOP is proposed for extracting complex behaviors, such as the refusal to answer harmful queries, from Large Language Models (LLMs). This method…
Simulation-based inference for rapid Bayesian parameter estimation in epidemiological models: a comparison with MCMC
Researchers compare Simulation-Based Inference (SBI) with Markov chain Monte Carlo (MCMC) methods for Bayesian calibration of a mechanistic epidemiological model using COVID-19 ICU occupancy data from Germany in 2020…
