Signals 4 · free daily AI digest

OPEN-1B: A Fully Auditable Training Run

Paper recorded by Signals 4 on 2026-09-15 in cs.LG. Abstract reproduced from arXiv; link to the original below.

Published 2026-09-15 on arXiv · recorded by Signals 4 on 2026-09-16

Category: cs.LG · 机器学习 · first seen 2026-09-16

Abstract

Open-source language models have a reproducibility problem. Despite releasing weights, training data, and recipes, none of them are provably reproducible due to the non-associativity of floating-point arithmetic. Deep learning frameworks often offer a deterministic execution mode, allowing reproducible operations on the same machines. Unfortunately, this determinism does not carry across hardware

Read on arXiv →

#33 most recent of 215 cs.LG papers we have recorded · ↑ newer: Bridging the Confidence Gap: Temperature Scaling for Calibrating Test- · ↓ older: Large Language Models Develop Belief State Geometry In-Context
Cite this page: OPEN-1B: A Fully Auditable Training Run: the #33 most recent of 215 cs.LG papers we have recorded (as of 2026-09-15). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/open-1b-a-fully-auditable-training-run.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable: papers.json
Related: More cs.LG papers · arXiv signals · All papers · Today in AI
Get 4 AI signals a day by email — free.
Subscribe free → See all plans →
Get 4 AI signals a day by email — free
All models · All repos · By company · Daily editions