mlsanityData to 14 Oct 2026

gpt-oss-120b & gpt-oss-20b Model Card

arXiv 2508.10925 ↗First posted 8 Aug 2025cs.CL cs.AIcited by 1,386

On the map: LLM reasoning → Test-time compute and efficient thinking

Abstract (arXiv)

We present gpt-oss-120b and gpt-oss-20b, two open-weight reasoning models that push the frontier of accuracy and inference cost. The models use an efficient mixture-of-expert transformer architecture and are trained using large-scale distillation and reinforcement learning. We optimize the models to have strong agentic capabilities (deep research browsing, python tool use, and support for developer-provided functions), all while using a rendered chat format that enables clear instruction following and role delineation. Both models achieve strong results on benchmarks ranging from mathematics, coding, and safety. We release the model weights, inference implementations, tool environments, and tokenizers under an Apache 2.0 license to enable broad use and further research.

Key facts. Cited by 1,386 papers in the map (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 28 of its 33 references that are in the map. First posted on arXiv 8 Aug 2025.

Builds on

28 of 33 references.

References · 28

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  5. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,703
  6. 2024GPT-4o System Cardcited by 4,588
  7. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,673
  8. 2017Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layercited by 3,509
  9. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,879
  10. 2023Efficient Streaming Language Models with Attention Sinkscited by 2,186
  11. 2020GLU Variants Improve Transformercited by 1,804
  12. 2020GShard: Scaling Giant Models with Conditional Computation and Automatic Shardingcited by 1,631
16 more
  1. 2019Generating Long Sequences with Sparse Transformerscited by 1,546
  2. 2019Root Mean Square Layer Normalizationcited by 1,480
  3. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domainscited by 1,402
  4. 2023GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpointscited by 1,346
  5. 2020On Layer Normalization in the Transformer Architecturecited by 922
  6. 2021GLaM: Efficient Scaling of Language Models with Mixture-of-Expertscited by 897
  7. 2019Fast Transformer Decoding: One Write-Head is All You Needcited by 717
  8. 2021BBQ: A Hand-Built Bias Benchmark for Question Answeringcited by 701
  9. 2023YaRN: Efficient Context Window Extension of Large Language Modelscited by 601
  10. 2024The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructionscited by 493
  11. 2025Humanity's Last Examcited by 491
  12. 2025HealthBench: Evaluating Large Language Models Towards Improved Human Healthcited by 429
  13. 2024A StrongREJECT for Empty Jailbreakscited by 385
  14. 2025PaperBench: Evaluating AI's Ability to Replicate AI Researchcited by 321
  15. 2024Deliberative Alignment: Reasoning Enables Safer Language Modelscited by 308
  16. 2024LAB-Bench: Measuring Capabilities of Language Models for Biology Researchcited by 156

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 276; 2026: 1110.

1,386 papers.

Sort:

By area:

  1. 2026DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligencecited by 1,040
  2. 2025Kimi Linear: An Expressive, Efficient Attention Architecturecited by 196
  3. 2026MiMo-V2-Flash Technical Reportcited by 191
  4. 2025A Survey of Reinforcement Learning for Large Reasoning Modelscited by 180
  5. 2026Cosmos 3: Omnimodal World Models for Physical AIcited by 165
  6. 2026Beyond Benchmarks: MathArena as an Evaluation Platform for Mathematics with LLMscited by 137
  7. 2026DFlash: Block Diffusion for Flash Speculative Decodingcited by 135
  8. 2025CWM: An Open-Weights LLM for Research on Code Generation with World Modelscited by 93
  9. 2025MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scalingcited by 81
  10. 2026MOPD: Multi-Teacher On-Policy Distillation for Capability Integration in LLM Post-Trainingcited by 72
  11. 2026Qwen3-Coder-Next Technical Reportcited by 71
  12. 2025Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Taskscited by 66
38 more
  1. 2026PersonaPlex: Voice and Role Control for Full Duplex Conversational Speech Modelscited by 63
  2. 2025End-to-End Test-Time Training for Long Contextcited by 62
  3. 2026Rethinking Rubric Generation for Improving LLM Judge and Reward Modeling for Open-ended Taskscited by 62
  4. 2025Hilbert: Recursively Building Formal Proofs with Informal Reasoningcited by 60
  5. 2025Baichuan-M2: Scaling Medical Capability with Large Verifier Systemcited by 59
  6. 2025SWE-EVO: Benchmarking Coding Agents in Long-Horizon Software Evolution Scenarioscited by 59
  7. 2025Towards General Agentic Intelligence via Environment Scalingcited by 54
  8. 2025Attention Sinks and Compression Valleys in LLMs are Two Sides of the Same Coincited by 52
  9. 2026SAGE: Scalable Agentic 3D Scene Generation for Embodied AIcited by 50
  10. 2025Nemotron-Cascade: Scaling Cascaded Reinforcement Learning for General-Purpose Reasoning Modelscited by 48
  11. 2025WebWeaver: Structuring Web-Scale Evidence with Dynamic Outlines for Open-Ended Deep Researchcited by 46
  12. 2025Continuum: Efficient and Robust Multi-Turn LLM Agent Scheduling with KV Cache Time-to-Livecited by 44
  13. 2026MOVA: Towards Scalable and Synchronized Video-Audio Generationcited by 43
  14. 2026Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoningcited by 42
  15. 2025Kimi-Dev: Agentless Training as Skill Prior for SWE-Agentscited by 41
  16. 2026Lost in Simulation: LLM-Simulated Users are Unreliable Proxies for Human Users in Agentic Evaluationscited by 40
  17. 2026Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thoughtcited by 40
  18. 2025ParaThinker: Native Parallel Thinking as a New Paradigm to Scale LLM Test-time Computecited by 38
  19. 2026Nemotron 3 Super: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoningcited by 38
  20. 2025Sycophancy Is Not One Thing: Causal Separation of Sycophantic Behaviors in LLMscited by 37
  21. 2026OpenResearcher: A Fully Open Pipeline for Long-Horizon Deep Research Trajectory Synthesiscited by 37
  22. 2025VCRL: Variance-based Curriculum Reinforcement Learning for Large Language Modelscited by 36
  23. 2026On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stabilitycited by 35
  24. 2025SFR-DeepResearch: Towards Effective Reinforcement Learning for Autonomously Reasoning Single Agentscited by 31
  25. 2025ChatInject: Abusing Chat Templates for Prompt Injection in LLM Agentscited by 31
  26. 2026IndexCache: Accelerating Sparse Attention via Cross-Layer Index Reusecited by 31
  27. 2025Scaling Behaviors of LLM Reinforcement Learning Post-Training: An Empirical Study in Mathematical Reasoningcited by 29
  28. 2025Why Do LLM Agents Fail in Exploring New Environments? A World-Modeling Perspectivecited by 28
  29. 2025Nemotron-Math: Efficient Long-Context Distillation of Mathematical Reasoning from Multi-Mode Supervisioncited by 26
  30. 2025What Characterizes Effective Reasoning? Revisiting Length, Review, and Structure of CoTcited by 25
  31. 2026Reasoning Models Struggle to Control their Chains of Thoughtcited by 25
  32. 2025TritonRL: Training LLMs to Think and Code Triton Without Cheatingcited by 24
  33. 2025BrowseSafe: Understanding and Preventing Prompt Injection Within AI Browser Agentscited by 24
  34. 2026Rethinking Failure Attribution in Multi-Agent Systems: A Multi-Perspective Benchmark and Evaluationcited by 24
  35. 2025A Survey on Agentic Multimodal Large Language Modelscited by 23
  36. 2026A Unified View of Attention and Residual Sinks: Outlier-Driven Rescaling is Essential for Transformer Trainingcited by 23
  37. 2026CLI-Gym: Scalable CLI Task Generation via Agentic Environment Inversioncited by 23
  38. 2025SwiReasoning: Switch-Thinking in Latent and Explicit for Pareto-Superior Reasoning LLMscited by 22

50 most cited of 1,386 All 1,386 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Advancing General-Purpose Reasoning Models with Modular Gradient Surgery
  2. 2024Marco-o1: Towards Open Reasoning Models for Open-Ended Solutions
  3. 2025AlphaApollo: A System for Deep Agentic Reasoning
  4. 2025Reasoning Vectors: Transferring Chain-of-Thought Capabilities via Task Arithmetic
  5. 2024OpenR: An Open Source Framework for Advanced Reasoning with Large Language Models
  6. 2025ShorterBetter: Guiding Reasoning Models to Find Optimal Inference Length for Efficient Reasoning
  7. 2025Marco-o1 v2: Towards Widening The Distillation Bottleneck for Reasoning Models
  8. 2025CoreThink: A Symbolic Reasoning Layer to reason over Long Horizon Tasks with LLMs