Logics-SWE-Qwen3.6-27B

Model card for Logics-SWE-Qwen3.6-27B. The technical report is in preparation.

News

• [2026.09.18] Released Logics-SWE-Qwen3.6-27B under the Apache-2.0 license.
• The technical report is in preparation. A link will be added when available.

Overview

Logics-SWE-Qwen3.6-27B is a 27B-parameter model developed for repository-level software engineering agents. Starting from the Qwen3.6-27B model used in our study, it combines category-aware expert development with multi-teacher on-policy distillation into a single deployment policy.

Repository-level tasks require agents to navigate code, edit files, execute commands, inspect feedback, and iteratively repair their solutions. Our work starts from the category see-saw: aggregate progress during joint RL can conceal opposing changes across task categories. Category splitting alone does not guarantee stronger experts. We therefore develop the experts through RRE before integrating their learned behaviors into one model.

Stage 1: Refresh–Repair–Expand expert development

SWE Labeler organizes training tasks into three repository-domain categories:

• A: service, data, and security.
• B: user-facing applications.
• C: systems, tooling, and runtimes.

All experts start from the same base model. Refresh–Repair–Expand (RRE) alternates executable-reward RL with supervised replay of each expert's own verified successful trajectories. Mastery refreshes track how task success rates change; repair consolidates successful behavior, and expansion revisits additional tasks to identify the next training frontier. Agentic-miniRL provides the shared long-horizon RL recipe.

Stage 2: Single-policy integration with MOPD

The student starts from the common base and generates its own trajectories. Each training task is routed to its corresponding category expert for token-level supervision. Reference-anchored extrapolation augments imitation, using a positive teacher–reference gap gate. The student receives no direct environment-reward term during this pure distillation stage.

Neither expert development nor integration uses an external model to provide solution trajectories or action targets.

Here, integration means on-policy distillation, not arithmetic averaging of expert weights. Inference uses one model; category labels and separate expert models are not required for deployment.

Intended use: research on repository-level issue resolution, coding agents, long-horizon post-training, expert development, and policy distillation.

Experimental Results

Values are mean task resolution (%) ± population standard deviation across three evaluation rounds, with one candidate patch per task per round.

Compared with the base model, Logics-SWE-Qwen3.6-27B improves mean task resolution from 52.64% to 58.04% on Pro-618 (+5.40 percentage points) and from 56.22% to 59.00% on SWE-bench Multilingual (+2.78 percentage points).

Evaluation scope

• Pro-618 retains 618 of 731 SWE-bench Pro tasks using documented public audit findings. Its A/B/C counts are 221/201/196. These are subset results, not full SWE-bench Pro scores.
• SWE-bench Multilingual includes all 300 tasks. A/B/C contain 72/25/176 tasks; 27 unrouted tasks remain in the Full denominator.
• Scores measure the complete agent setup, not standalone text generation. Pro uses the study's R2E-Gym-based setup; Multilingual uses a SWE-agent setup. Absolute scores across the two scaffolds are not directly comparable.
• Unsuccessful or missing evaluations remain in the fixed task denominator.

Quickstart

Transformers: text-generation smoke test

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "Logics-MLLM/Logics-SWE-Qwen3.6-27B"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    dtype=torch.bfloat16,
    device_map="auto",
)
model.eval()

messages = [{
    "role": "user",
    "content": "A parser crashes on an empty input file. Describe how you would "
               "investigate the cause, implement a fix, and test for regressions.",
}]
inputs = tokenizer.apply_chat_template(
    messages,
    tokenize=True,
    add_generation_prompt=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device)

with torch.inference_mode():
    output = model.generate(
        **inputs,
        max_new_tokens=1024,
        do_sample=True,
        temperature=1.0,
        top_p=0.95,
        top_k=20,
    )
completion = output[0, inputs["input_ids"].shape[-1]:]
print(tokenizer.decode(completion, skip_special_tokens=True))

The example follows the Transformers chat-template interface. Its short generation budget is for demonstration only. Repository-level agent use additionally requires the matching system prompt, tool schema, action parser, sandbox, and verifier. Model-generated commands should not be executed automatically outside an isolated, permission-controlled environment.

Training Data and Release Scope

Alongside the released model weights, we plan to make the following resources publicly available:

• A curated subset of instance-level training data, subject to approval.
• The label taxonomy, annotation guidelines, and SWE Labeler code.
• Anonymous training manifests, benchmark outcomes, aggregate results, and training and evaluation configuration specifications.

Citation

The technical report is in preparation. Citation information will be added upon release.

Acknowledgements

We thank the Qwen, ROLL, R2E-Gym, SWE-agent, SWE-bench, SWE-bench Pro, and SWE-bench Multilingual contributors, and the research communities behind MiniRL, MOPD, and ExOPD.

License

The model weights are released under the Apache License 2.0.