On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,657 papers · 148 categories

Trend · papers per month

count % of corpus

4 results for “IVON”

Variational Low-Rank Adaptation Using IVONcs.LG

Improved IVON boosts LoRA model accuracy and calibration.

problem Improving the accuracy and calibration of large language models.

method Replaced AdamW with IVON for finetuning Llama-2.

result IVON improves Llama-2 accuracy by 2.8% and calibration error by 4.6%.

Large Language Models LoRA IVON Bayesian Accuracy

2024-11-07abs ↗pdf ↗

Variational Learning is Effective for Large Deep Networkscs.LG

IVON optimizes large neural networks, matching or outperforming Adam.

problem The inefficacy of variational learning in large neural networks.

method Improved Variational Online Newton (IVON) optimizer.

result IVON consistently matches or outperforms Adam for large networks.

Variational Learning Large Neural Networks IVON Optimizer Adam Predictive Uncertainty

2024-02-27abs ↗pdf ↗

Improving LoRA with Variational Learningcs.LG

IVON improves LoRA finetuning with minimal overhead and significant accuracy gains.

problem Improving LoRA finetuning with Bayesian methods.

method IVON, a variational algorithm, with posterior pruning.

result Significant accuracy improvements over AdamW and other Bayesian methods.

LoRA Bayesian methods IVON Variational learning Accuracy improvement

2025-06-17abs ↗pdf ↗

A Bayesian Interpretation of Adaptive Low-Rank Adaptationcs.LG

Bayesian approach improves AdaLoRA's performance and efficiency.

problem Improving the efficiency and performance of adaptive low-rank adaptation.

method Utilized Bayesian metrics and the Improved Variational Online Newton (IVON) optimizer for adaptive parameter budget allocation.

result Bayesian counterpart outperforms sensitivity-based importance metric and is faster than AdaLoRA.

Bayesian interpretation AdaLoRA adaptive low-rank adaptation signal-to-noise ratio Improved Variational Online Newton

2024-09-16abs ↗pdf ↗