Atlas / Reports
Reports
Technical reports are treated here as public evidence trails: a way to connect names, organizations, and moments in the LLM timeline.
GLM-5: Thinking, Coding, and Agentic Intelligence
Z.ai
Large Language Models · 2602.15763 · 2026-02-17
CWM: An Open-Weights LLM for Research on Code Generation with World Models
Meta AI
Code Language Models · 2509.12054 · 2025-09-24
Apple Intelligence Foundation Language Models: Tech Report 2025
Apple
Multimodal Language Models · 2507.13575 · 2025-07-16
Magistral: Efficient Training of Small Language Models for Reasoning
Mistral AI
Reasoning Models · 2506.10910 · 2025-06-12
Amazon Nova Sonic Technical Report
Amazon
Speech Language Models · 2505.11298 · 2025-05-15
Phi-4-mini-reasoning: Exploring the Limits of Small Reasoning Language Models in Math
Microsoft
Reasoning Models · 2504.21233 · 2025-04-29
Phi-4-reasoning Technical Report
Microsoft
Reasoning Models · 2504.21318 · 2025-04-29
Hunyuan-T1: Scaling Up Test-Time Compute with Open-Source Reinforcement Learning
Tencent Hunyuan
Reasoning Models · 2504.02234 · 2025-04-03
Command A: An Enterprise-Ready Large Language Model
Cohere
Large Language Models · 2504.00698 · 2025-04-01
Mistral Small 3.1 Technical Report
Mistral AI
Large Language Models · 2503.23335 · 2025-03-31
QwQ-32B: Embracing the Power of Reinforcement Learning
Qwen
Reasoning Models · 2503.20735 · 2025-03-27
Qwen2.5-Omni Technical Report
Qwen
Multimodal Models · 2503.20215 · 2025-03-23
Phi-4 Technical Report
Microsoft
Language Models · 2503.01743 · 2025-03-03
Qwen2.5-VL Technical Report
Qwen
Vision-Language Models · 2502.13923 · 2025-02-19
Kimi k1.5: Scaling Reinforcement Learning with LLMs
Moonshot AI
Large Language Models · 2501.12599 · 2025-01-21
2 OLMo 2 Furious
Ai2
Large Language Models · 2501.00656 · 2024-12-31
DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding
DeepSeek
Vision-Language Models · 2412.10302 · 2024-12-12
NVLM: Open Frontier-Class Multimodal LLMs
NVIDIA
Multimodal Language Models · 2412.04468 · 2024-12-05
GLM-4-Voice: Towards Intelligent and Human-Like End-to-End Spoken Chatbots
Z.ai
Audio Language Models · 2412.02612 · 2024-12-04
Tulu 3: Pushing Frontiers in Open Language Model Post-Training
Ai2
Large Language Models · 2411.15124 · 2024-11-22
JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation
DeepSeek
Vision-Language Models · 2411.07975 · 2024-11-11
Hunyuan-Large: An Open-Source MoE Model with 52 Billion Activated Parameters by Tencent
Tencent Hunyuan
Large Language Models · 2411.02265 · 2024-11-04
Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation
DeepSeek
Vision-Language Models · 2410.13848 · 2024-10-18
Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models
Ai2
Vision-Language Models · 2409.17146 · 2024-09-25
Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
Qwen
Vision-Language Models · 2409.12191 · 2024-09-18
OLMoE: Open Mixture-of-Experts Language Models
Ai2
Large Language Models · 2409.02060 · 2024-09-03
DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search
DeepSeek
Mathematical Reasoning Models · 2408.08152 · 2024-08-14
Apple Intelligence Foundation Language Models
Apple
Multimodal Language Models · 2407.21075 · 2024-07-29
Qwen2-Audio Technical Report
Qwen
Audio Language Models · 2407.10759 · 2024-07-14
Open Instruct: A Simple Method for Aligning Language Models with Human Preferences
Ai2
Large Language Models · 2406.18405 · 2024-06-26
DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence
DeepSeek
Code Language Models · 2406.11931 · 2024-06-17
Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
Microsoft
Language Models · 2404.14219 · 2024-04-22
Jamba: A Hybrid Transformer-Mamba Language Model
AI21 Labs
Language Models · 2403.19887 · 2024-03-28
DeepSeek-VL: Towards Real-World Vision-Language Understanding
DeepSeek
Vision-Language Models · 2403.05525 · 2024-03-08
Nemotron-4 15B Technical Report
NVIDIA
Large Language Models · 2402.16819 · 2024-02-26
Many-shot Jailbreaking
Anthropic
Alignment and Safety · 2402.03206 · 2024-02-12
SPIrit-LM: Interleaved Spoken and Written Language Model
Meta AI
Speech Language Models · 2402.05755 · 2024-02-09
DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
DeepSeek
Mathematical Reasoning Models · 2402.03300 · 2024-02-06
Mixtral of Experts
Mistral AI
Large Language Models · 2401.04088 · 2024-01-08
Tulu 2: Demystifying the Effectiveness of RLHF and Reinforcement Learning with Human Feedback
Ai2
Large Language Models · 2311.10702 · 2023-11-17
Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models
Qwen
Audio Language Models · 2311.07919 · 2023-11-13
Mistral 7B
Mistral AI
Large Language Models · 2310.06825 · 2023-10-10
Collective Constitutional AI: Aligning a Language Model with Public Input
Anthropic
Alignment and RLHF · 2310.01835 · 2023-10-03
Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
Qwen
Vision-Language Models · 2308.12966 · 2023-08-24
Code Llama: Open Foundation Models for Code
Meta AI
Code Language Models · 2308.12950 · 2023-08-24
LLaMA: Open and Efficient Foundation Language Models
Meta AI
Large Language Models · 2302.13971 · 2023-02-27
Constitutional AI: Harmlessness from AI Feedback
Anthropic
Alignment and RLHF · 2212.08073 · 2022-12-15
Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
Anthropic
Alignment and RLHF · 2204.05862 · 2022-04-12
Training language models to follow instructions with human feedback
OpenAI
Alignment · 2203.02155 · 2022-03-04
Language Models are Few-Shot Learners
OpenAI
Large Language Models · 2005.14165 · 2020-05-28
MiniMax-Text-01
MiniMax
Large Language Models · 2501.08338
MiniMax-VL-01
MiniMax
Vision-Language Models · 2501.08336
Tracing the thoughts of a large language model
Anthropic
Interpretability · 2503.21435
On the Biology of a Large Language Model
Anthropic
Interpretability · 2504.19173
Constitutional Classifiers++: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming
Anthropic
Alignment and Safety · 2601.04603
Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming
Anthropic
Alignment and Safety · 2501.18837
Voxtral Technical Report
Mistral AI
Speech Language Models · 2507.13264
Nemotron-CrossThink: Efficient Knowledge Distillation of Long Chain-of-Thought Reasoning
NVIDIA
Reasoning Models · 2504.13941
Nemotron 3 Super: Open, efficient mixture-of-experts hybrid mamba-transformer model for agentic reasoning
NVIDIA
Reasoning Models · 2601.11868
NVIDIA Nemotron 3: Efficient and Open Intelligence
NVIDIA
Large Language Models · 2512.20856
Nemotron 3 nano: Open, efficient mixture-of-experts hybrid mamba-transformer model for agentic reasoning
NVIDIA
Reasoning Models · 2512.20848
NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Model
NVIDIA
Reasoning Models · 2508.14444
Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models
NVIDIA
Large Language Models · 2504.03624
DeepSeek-Prover-V2: Advancing Formal Mathematical Reasoning via Reinforcement Learning and Monte-Carlo Tree Search with Proof Assistant Feedback
DeepSeek
Mathematical Reasoning Models · 2508.03613
Large Concept Models: Language Modeling in a Sentence Representation Space
Meta AI
Language Models · 2502.06018
Qwen3-Omni Technical Report
Qwen
Multimodal Models · 2509.17765
MiniMax-Speech: Intrinsic Zero-Shot Speech Understanding for Advanced Foundation Models
MiniMax
Speech Language Models · 2505.07916
Magma: A Foundation Model for Multimodal AI Agents
Microsoft
Multimodal Agent Models · 2502.13130
GLM-4.5: Agentic, Reasoning, and Coding Foundation Models
Z.ai
Language Models · 2508.06471
GLM-Z1-Rumination: An Open Frontier-Class Reasoning Model Through Test-Time Scaling
Z.ai
Reasoning Models · 2506.17434
Auditing language models for hidden objectives
Anthropic
Alignment and Safety · 2507.11473
Alignment faking in large language models
Anthropic
Alignment and Safety · 2412.14093
Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
Anthropic
Alignment and Safety · 2401.05566
Amazon Nova Premier Technical Report
Amazon
Large Language Models · 2504.01081
Aya Vision: Advancing the Frontier of Multilingual Multimodality
Cohere
Multimodal Language Models · 2410.14756
MM1.5: Methods, Analysis and Insights from Multimodal LLM Fine-tuning
Apple
Multimodal Language Models · 2409.20566
MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training
Apple
Multimodal Language Models · 2403.09611
OpenAI o3 and o4-mini System Card
OpenAI
Reasoning Models · 2504.21798
OpenAI o1 System Card
OpenAI
Reasoning Models · 2412.16720
Nemotron-4 340B Technical Report
NVIDIA
Large Language Models · 2406.11704
Jamba 1.5 Technical Report
AI21 Labs
Language Models · 2508.15167
Qwen2.5-Coder Technical Report
Qwen
Code Language Models · 2409.12186
OLMo: Accelerating the Science of Language Models
Ai2
Large Language Models · 2402.00838
Pixtral 12B
Mistral AI
Multimodal Large Language Models · 2410.17897
Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling
DeepSeek
Multimodal Large Language Models · 2501.17811
MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention
MiniMax
Reasoning Large Language Models · 2506.13585
Chameleon: Mixed-Modal Early-Fusion Foundation Models
Meta AI
Multimodal Large Language Models · 2405.09818
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
DeepSeek
Large Language Models · 2501.12948
GLM-4.1V-Thinking and GLM-4.5V: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
Z.ai
Multimodal Models · 2507.01006
MiniMax-01: Scaling Foundation Models with Lightning Attention
MiniMax
Large Language Models · 2501.08313
The Llama 3 Herd of Models
Meta AI
Large Language Models · 2407.21783
Llama 2: Open Foundation and Fine-Tuned Chat Models
Meta AI
Large Language Models · 2307.09288
Kimi K2.5: Visual Agentic Intelligence
Moonshot AI
Multimodal Agentic Models · 2602.02276
Kimi-VL Technical Report
Moonshot AI
Vision-Language Models · 2504.07491
DeepSeek LLM Technical Report
DeepSeek
Large Language Models · 2401.02954
DeepSeek-V2 Technical Report
DeepSeek
Large Language Models · 2405.04434
DeepSeek-V3 Technical Report
DeepSeek
Large Language Models · 2412.19437
Qwen2.5 Technical Report
Qwen
Large Language Models · 2412.15115
Qwen3 Technical Report
Qwen
Large Language Models · 2505.09388
Qwen Technical Report
Qwen
Large Language Models · 2309.16609
GPT-4 Technical Report
OpenAI
Large Language Models · 2303.08774