PulseAugur
EN
LIVE 09:01:37

New Loki method improves pretrained models without retraining

By PulseAugur Editorial · [1 sources] ·

Researchers have developed a new method called Loki that adapts pretrained machine learning models to improve their performance on new classes without additional training. This technique replaces the standard prediction rule with the Fréchet mean, leveraging metric information within label spaces. Loki has demonstrated significant gains, including up to a 29.7% relative improvement over SimCLR on ImageNet and a 10.5% improvement on pretrained zero-shot models like CLIP when external metrics are unavailable. AI

IMPACT This method could enhance the efficiency of adapting large models to new tasks, reducing the need for extensive retraining.

RANK_REASON The cluster contains an academic paper detailing a new method for adapting pretrained models. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

New Loki method improves pretrained models without retraining

How we ranked this

Signal score
15 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The cluster contains an academic paper detailing a new method for adapting pretrained models. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. arXiv cs.AI TIER_1 English(EN) · Nicholas Roberts, Xintong Li, Dyah Adila, Sonia Cromp, Tzu-Heng Huang, Jitian Zhao, Frederic Sala ·

    Geometry-Aware Adaptation for Pretrained Models

    arXiv:2307.12226v3 Announce Type: replace-cross Abstract: Machine learning models -- including prominent zero-shot models -- are often trained on datasets whose labels are only a small proportion of a larger label space. Such spaces are commonly equipped with a metric that relate…

RELATED ENTITIES

RELATED TOPICS