Composer 2 Technical Report
Composer 2 is a specialized coding model trained through phased learning approaches that demonstrates superior performance on real-world software engineering tasks. (0 upvotes on HuggingFace)
Published on Mar 25
Authors:
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
Abstract
Composer 2 is a specialized coding model trained through phased learning approaches that demonstrates superior performance on real-world software engineering tasks.
AI-generated summary
Composer 2 is a specialized model designed for agentic software engineering. The model demonstrates strong long-term planning and coding intelligence while maintaining the ability to efficiently solve problems for interactive use. The model is trained in two phases: first, continued pretraining to improve the model's knowledge and latent coding ability, followed by large-scale reinforcement learning to improve end-to-end coding performance through stronger reasoning, accurate multi-step execution, and coherence on long-horizon realistic coding problems. We develop infrastructure to support training in the same Cursor harness that is used by the deployed model, with equivalent tools and structure, and use environments that match real problems closely. To measure the ability of the model on increasingly difficult tasks, we introduce a benchmark derived from real software engineering problems in large codebases including our own. Composer 2 is a frontier-level coding model and demonstrates a process for training strong domain-specialized models. On our CursorBench evaluations the model achieves a major improvement in accuracy compared to previous Composer models (61.3). On public benchmarks the model scores 61.7 on Terminal-Bench and 73.7 on SWE-bench Multilingual in our harness, comparable to state-of-the-art systems.
View arXiv page View PDF Add to collection
Get this paper in your agent:
hf papers read 2603.24477
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2603.24477 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2603.24477 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2603.24477 in a Space README.md to link it from this page.
Collections including this paper 2
Sign in to highlight and annotate this article
Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
researchpaperarxiv
With an eye on China, Japan looks to kamikaze drones and low-cost missiles
Japan plans to introduce a fleet of kamikaze drones and low-cost missiles to boost deterrence against regional threats including China, according to Japanese media reports. The Yomiuri newspaper and Kyodo news agency reported on Wednesday that the strategy was focused on “integrated attacks” from unmanned aerial vehicles and long-range stand-off missiles, citing government and ruling coalition sources. They said the drones and missiles would be used to break down enemy air defences and...
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Research Papers
Realistic Lip Motion Generation Based on 3D Dynamic Viseme and Coarticulation Modeling for Human-Robot Interaction
arXiv:2604.01756v1 Announce Type: new Abstract: Realistic lip synchronization is essential for the natural human-robot non-verbal interaction of humanoid robots. Motivated by this need, this paper presents a lip motion generation framework based on 3D dynamic viseme and coarticulation modeling. By analyzing Chinese pronunciation theory, a 3D dynamic viseme library is constructed based on the ARKit standard, which offers coherent prior trajectories of lips. To resolve motion conflicts within continuous speech streams, a coarticulation mechanism is developed by incorporating initial-final (Shengmu-Yunmu) decoupling and energy modulation. After developing a strategy to retarget high-dimensional spatial lip motion to a 14-DOF lip actuation system of a humanoid head platform, the efficiency and
3-D Relative Localization for Multi-Robot Systems with Angle and Self-Displacement Measurements
arXiv:2604.01703v1 Announce Type: new Abstract: Realizing relative localization by leveraging inter-robot local measurements is a challenging problem, especially in the presence of measurement noise. Motivated by this challenge, in this paper we propose a novel and systematic 3-D relative localization framework based on inter-robot interior angle and self-displacement measurements. Initially, we propose a linear relative localization theory comprising a distributed linear relative localization algorithm and sufficient conditions for localizability. According to this theory, robots can determine their neighbors' relative positions and orientations in a purely linear manner. Subsequently, in order to deal with measurement noise, we present an advanced Maximum a Posterior (MAP) estimator by a
Coupler Position Optimization and Channel Estimation for Flexible Coupler Antenna Aided Multiuser Communication
arXiv:2602.11319v2 Announce Type: replace-cross Abstract: In this paper, we propose a distributed flexible coupler antenna (FCA) array to enhance communication performance with low hardware cost. At each FCA, there is one fixed-position active antenna and multiple passive couplers that can move within a designated region around the active antenna. Moreover, each FCA is equipped with a local processing unit (LPU). All LPUs exchange signals with a central processing unit (CPU) for joint signal processing. We study an FCA-aided multiuser multiple-input multiple-output (MIMO) system, where an FCA array base station (BS) is deployed to enhance the downlink communication between the BS and multiple single-antenna users. We formulate optimization problems to maximize the achievable sum rate of us
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!