Sasha Rush
ML researcher at Cursor
Makes model research inspectable through annotated code and training reports, from a Transformer implementation to the training and evaluation of a coding model.
Why read Sasha Rush?
Rush offers a route from using a coding model to understanding how one is trained. His Composer 2 introduction explains continued pretraining followed by reinforcement learning in coding environments. The coauthored technical report adds experiments and explains why the team's evaluation includes ambiguous requests and changes across several files.
For the underlying architecture, read his Annotated Transformer. It connects the original paper's equations to a line-by-line implementation, then demonstrates a copy task and translation. Some code dates from 2018; the page links an updated implementation. Familiarity with Python, tensors and basic neural networks will help.
Selected work
-
A technical report on Composer 2
How is a coding model trained for an agent's work? Rush introduces continued pretraining, reinforcement learning in Cursor sessions and evaluation on tasks from the team's own engineering work. The article leads to the full team report.
Cite this selection
XFollowList, “A technical report on Composer 2” reading notes in the Sasha Rush profile. https://xfollowlist.com/people/sasha-rush#work-24bfc589cda3
The linked work is the original source; this permalink identifies XFollowList’s reason to read it. Original work published .
Profile-source checking date and references · Dated reading map
-
Composer 2 Technical Report
What evidence supports a training change? This coauthored Cursor Research report connects the recipe to ablations, agent behavior and the design of CursorBench. Read the benchmark conditions alongside the team's reported results.
Cite this selection
XFollowList, “Composer 2 Technical Report” reading notes in the Sasha Rush profile. https://xfollowlist.com/people/sasha-rush#work-a8b5d47c724a
The linked work is the original source; this permalink identifies XFollowList’s reason to read it. Original work published .
Profile-source checking date and references · Dated reading map
-
The Annotated Transformer
How do the Transformer's equations become code? Rush's annotated implementation of Vaswani and coauthors' architecture covers attention, masking, training and translation, with help credited to Vincent Nguyen and Guillaume Klein.
Cite this selection
XFollowList, “The Annotated Transformer” reading notes in the Sasha Rush profile. https://xfollowlist.com/people/sasha-rush#work-91962865e154
The linked work is the original source; this permalink identifies XFollowList’s reason to read it. Original work published .
Profile-source checking date and references · Dated reading map
-
Llama2.rs
What does a small local language-model program do? Rush and Rachit Singh extend a Rust port of llama2.c. The README exposes model conversion, memory-mapped weights and decoder settings; some Rust familiarity is useful.
Cite this selection
XFollowList, “Llama2.rs” reading notes in the Sasha Rush profile. https://xfollowlist.com/people/sasha-rush#work-986d92ab2c52
The linked work is the original source; this permalink identifies XFollowList’s reason to read it. Original publication date is unknown.
Profile-source checking date and references · Dated reading map
Projects & roles
The Annotated Transformer
An annotated implementation of the original Transformer architecture; Rush's original article links the project.
Composer 2 research
Cursor Research's coauthored account of training a model for agentic software engineering, including Rush among the authors.
Sources & further reading
Profile sources checked . This describes checking these references, separately from the publication dates of selected works.
- His current site identifies his Cursor research, lists his earlier Harvard, Cornell and Hugging Face roles as past, and links @srush_nlp.Link to this source note
- Cursor's original article carries Sasha Rush's dated byline and links the technical report.Link to this source note
- The author record includes Alexander M. Rush, the name used in his academic publications.Link to this source note
- Original material read for the local-model selection.Link to this source note
Cite this profile’s source check
XFollowList, Sasha Rush profile sources. Sources checked 2026-10-09. https://xfollowlist.com/people/sasha-rush#sources-checked