Papers › Branch-Train-MiX: Mixing Expert LLMs into a Mixture-of-Experts LLM
Branch-Train-MiX: Mixing Expert LLMs into a Mixture-of-Experts LLM
Sainbayar Sukhbaatar, Olga Golovneva, Vasu Sharma, Hu Xu, Xi Victoria Lin, Baptiste Rozière, Jacob Kahn, Daniel Li, Wen-tau Yih, Jason Weston, Xian Li
We investigate efficient methods for training Large Language Models (LLMs) to possess capabilities in multiple specialized domains, such as coding, math reasoning and world knowledge. Our method, named Branch-Train-MiX (BTX), starts from a seed model, which is branched to train experts in embarrassingly parallel fashion with high throughput and reduced communication cost. After individual experts are asynchronously trained, BTX brings together their feedforward parameters as experts in Mixture-of-Expert (MoE) layers and averages the remaining parameters, followed by an MoE-finetuning stage to learn token-level routing. BTX generalizes two special cases, the Branch-Train-Merge method, which does not have the MoE finetuning stage to learn routing, and sparse upcycling, which omits the stage of training experts asynchronously. Compared to alternative approaches, BTX achieves the best accuracy-efficiency tradeoff.
In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
Code
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Not run by Syntology. Nothing on this page verifies that the listed code works.
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Arithmetic Reasoning | GSM8K | Branch-Train-MiX 4x7B (sampling top-2 experts) | Accuracy | 37.1 | #141 of 164 | Archive leaderboard | report |
| Code Generation | MBPP | Branch-Train-Merge 4x7B (top-2) | Accuracy | 42.6 | #80 of 99 | Archive leaderboard | report |
| Code Generation | MBPP | Branch-Train-MiX 4x7B (sampling top-2 experts) | Accuracy | 39.4 | #84 of 99 | Archive leaderboard | report |
| Common Sense Reasoning | WinoGrande | Branch-Train-MiX 4x7B (sampling top-1 expert) | Accuracy | 70.6 | #38 of 77 | Archive leaderboard | report |
| Math Word Problem Solving | MATH | Branch-Train-MiX 4x7B (sampling top-2 experts) | Accuracy | 17.8 | #106 of 135 | Archive leaderboard | report |
| Multi-task Language Understanding | MML | Branch-Train-MiX 4x7B (sampling top-1 experts) | Average (%) | 53.2 | #24 of 44 | Archive leaderboard | report |
| Question Answering | TriviaQA | Branch-Train-MiX 4x7B (sampling top-2 experts) | EM | 57.1 | #40 of 56 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections