---
format: "aidr-story-markdown/v1"
id: "62e88035876bf5e5ce8b80aa7f6e4c5bf649f49858f7510bdea9b60a0abc61a8"
canonical_url: "https://aidr.today/62e88035?lang=en"
title: "A study of sequence weighting at scale"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-09-21T01:18:47.000Z"
category: "Models"
topics: ["anthropic","claude","multi-agent","open-source"]
source_urls: ["https://blog.janestreet.com/a-study-of-sequence-weighting-at-scale/"]
summary: "A study on sequence weighting across in-house and open-weight LMs found non-monotonic scaling behavior, where small to medium models learn data-specific patterns proportional to weights, while large models learn all patterns independent of weights."
---

# A study of sequence weighting at scale

> [Open the canonical story](<https://aidr.today/62e88035?lang=en>)

**Published:** 2026-09-21T01:18:47.000Z
**Category:** Models
**Topics:** anthropic, claude, multi\-agent, open\-source

## Summary

A study on sequence weighting across in\-house and open\-weight LMs found non\-monotonic scaling behavior, where small to medium models learn data\-specific patterns proportional to weights, while large models learn all patterns independent of weights\.

## Sources

- [Story source](<https://blog.janestreet.com/a-study-of-sequence-weighting-at-scale/>)

