Entity trail

Thomson-1.0-Small

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-26 / reddit-researcherThomson Reuters Released Thomson-1.0-Small, a Law and Tax Model Built on Qwen3.6-35B-A3BThomson Reuters published Thomson-1.0-Small on Hugging Face, a domain model for legal and tax work fine-tuned from Qwen3.6-35B-A3B, drawing 163 upvotes on r/LocalLLaMA. The name misleads on size, as several commenters noted expecting a 1B model. The most substantive pushback in the thread argues domain models should be trained to understand how the field works and then aggressively retrieve, rather than baking case and code knowledge into weights that go stale the moment the law changes.
2026-08-26 / thought-leaders-researcherThomson Reuters builds its own legal LLM on top of Qwen and open-weights the small versionThomson Reuters launched Thomson 1.0 on August 24, trained not from scratch but by continuing training from Alibaba's Qwen on decades of Westlaw and Practical Law case law, contracts, statutes and practitioner guidance, for a reported $40M and using under 10% of its legal corpus. The open-weight Thomson-1.0-Small on Hugging Face is a 35B-total/3B-active MoE built from Qwen3.6-35B-A3B, claiming 79.9% on Stanford LegalBench and 74.6% overall average, under a restrictive Polyform-Strict license. The pattern worth watching for builders is a data-rich incumbent skipping pretraining entirely and buying its differentiation with post-training on proprietary corpora.
2026-07-14 / hn-researcherThomson Reuters to Cut a 'Small Number' of Engineering JobsThomson Reuters said on July 13, 2026 it will cut a 'small number' of engineering roles, another data point in the ongoing reshaping of software-engineering headcount at large information/tech firms. Single-source (US News/Reuters wire headline); specific counts and full rationale not detailed in the available report.

Graph relationships

BENCHMARKED AGAINST
Stanford LegalBench
Thomson-1.0-Small claims 79.9% on Stanford LegalBench.
Source finding
RELEASED
Thomson Reuters
Thomson Reuters released Thomson-1.0-Small, a law and tax model fine-tuned from Qwen3.6-35B-A3B.
Source finding
RELEASED
Thomson Reuters -> Thomson-1.0-Small
Thomson Reuters released Thomson-1.0-Small, a law and tax model fine-tuned from Qwen3.6-35B-A3B.
Source finding

Source trail

Hugging Face (thomsonreuters/Thomson-1.0-Small)U.S. News / Reuters

Graph sources

entity graphfindings textkg edgeskg entities

Fetching from the wire…