AI Paper Atlas

Research map / Large language models / Agents, code and math

RedSage: A Cybersecurity Generalist LLM

Naufal Suryanto, Muzammal Naseer, Pengfei Li, Syed Talal Wasim, Jinhui Yi, Juergen Gall, Paolo Ceravolo, Ernesto Damiani

ICLR 2026 · Large language models › Agents, code and math · approach: LLMs for optimization and heuristics · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
continual pretraining on cybersecurity corpora (CFW then Seed)Qwen3-8B-Base · Cybersecurity benchmarks mean (5-shot)mean accuracy (%): 80.81 -> 84.56Improved
add CyberFineWeb CPT only (ablation)Qwen3-8B-Base · Cybersecurity benchmarks meanmean accuracy (%): 80.81 -> 82.66Improved
add RedSage-Seed CPT only (ablation)Qwen3-8B-Base · Cybersecurity benchmarks meanmean accuracy (%): 80.81 -> 84.45Improved
combine CFW then Seed CPT instead of Seed only (ablation)RedSage-8B · Cybersecurity benchmarks meanmean accuracy (%): 84.45 -> 84.56No clear effect
cybersecurity continual pretrainingQwen3-8B-Base · General benchmarks meanmean accuracy (%): 70.86 -> 69.23Hurt
SFT with RedSage-Conv and SmolTalk2 (Ins vs Qwen3-8B)Qwen3-8B · Cybersecurity benchmarks mean (0-shot)mean accuracy (%): 75.71 -> 81.30Improved
add DPO after SFT (ablation)RedSage-8B-Ins · Cybersecurity benchmarks mean (0-shot)mean accuracy (%): 81.30 -> 81.10No clear effect
add DPO after SFT (ablation)RedSage-8B-Ins · General benchmarks meanmean accuracy (%): 73.34 -> 74.33Improved
add DPO after SFT (ablation)RedSage-8B-Ins · IFEvalaccuracy (%): 79.97 -> 83.44Improved
add DPO after SFT (ablation)RedSage-8B-Ins · CTI-Bench RCMaccuracy (%): 76.70 -> 70.60Hurt
instruction tuning with RedSage data vs Qwen3-8BQwen3-8B · General benchmarks meanmean accuracy (%): 65.92 -> 74.33Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy