| Execution Guided Line-by-Line Code Generation | ✓ Link | 96.6 | EG-CFG (DeepSeek-V3-0324) | 2025-06-12 |
| SYMPHONY: Synergistic Multi-agent Planning with Heterogeneous Language Model Assembly | ✓ Link | 96.5 | SYMPHONY-L | 2026-01-30 |
| ThinkRouter: Efficient Reasoning via Routing Thinking between Latent and Discrete Spaces | | 96.49 | ThinkRouter (Qwen3-32B) | 2026-02-12 |
| SePO: Self-Evolving Prompt Agent for System Prompt Optimization | | 96.2 | SePO-Generalist | 2026-06-03 |
| AgentConductor: Topology Evolution for Multi-Agent Competition-Level Code Generation | | 95.1 | AgentConductor (3B) | 2026-02-19 |
| Introspective Diffusion Language Models | ✓ Link | 94.6 | I-DLM (32B) | 2026-04-13 |
| CopT: Contrastive On-Policy Thinking with Continuous Spaces for General and Agentic Reasoning | ✓ Link | 94.55 | CopT (Qwen3-8B) | 2026-05-19 |
| QualityFlow: An Agentic Workflow for Program Synthesis Controlled by LLM Quality Checks | | 94.2 | QualityFlow (Sonnet-3.5) | 2025-01-20 |
| MapCoder: Multi-Agent Code Generation for Competitive Problem Solving | ✓ Link | 93.2 | o1-mini + MapCoder (Hamming.ai) | 2024-05-18 |
| SYMPHONY: Synergistic Multi-agent Planning with Heterogeneous Language Model Assembly | ✓ Link | 92.7 | SYMPHONY-S | 2026-01-30 |
| From Code to Correctness: Closing the Last Mile of Code Generation with Hierarchical Debugging | ✓ Link | 92.4 | MGDebugger (DeepSeek-V3-0324) | 2024-10-02 |
| AgentCoder: Multi-Agent-based Code Generation with Iterative Testing and Optimisation | ✓ Link | 91.8 | GPT-4 + AgentCoder | 2023-12-20 |
| CODESIM: Multi-Agent Code Generation and Problem Solving through Simulation-Driven Planning and Debugging | ✓ Link | 90.7 | CodeSim (GPT4o) | 2025-02-08 |
| | 90.0 | Jiutian-大模型 | |
| Scalable Prompt Routing via Fine-Grained Latent Task Discovery | | 90 | FineRouter | 2026-03-19 |
| AgentCoder: Multi-Agent-based Code Generation with Iterative Testing and Optimisation | ✓ Link | 89.9 | GPT-3.5 Turbo (ChatGPT) + AgentCoder | 2023-12-20 |
| MapCoder: Multi-Agent Code Generation for Competitive Problem Solving | ✓ Link | 89.7 | MapCoder (GPT-4o) | 2024-05-18 |
| UCoder: Unsupervised Code Generation by Internal Probing of Large Language Models | | 89.7 | UCoder | 2025-12-19 |
| Loop the Loopies! | | 89.49 | Loopie-20B-A2B (Thinking) | 2026-07-17 |
| LLaDA2.0: Scaling Up Diffusion Language Models to 100B | ✓ Link | 88.29 | LLaDA2.0-flash | 2025-12-10 |
| Qwen2.5 Technical Report | ✓ Link | 88.2 | Qwen2.5-72B-Instruct | 2024-12-19 |
| Recursive Think-Answer Process for LLMs and VLMs | | 87.9 | Qwen-2.5-32B-R-TAP | 2026-03-02 |
| From Next-Token to Next-Block: A Principled Adaptation Path for Diffusion LLMs | ✓ Link | 87.6 | NBDiff-7B (Instruct) | 2025-12-07 |
| HyperCLOVA X THINK Technical Report | ✓ Link | 87.59 | HyperCLOVA X SEED 14B Think | 2025-06-27 |
| How Does Naming Affect LLMs on Code Analysis Tasks? | | 87.5 | GPT-4 (ChatGPT Plus) | 2023-07-24 |
| The Claude 3 Model Family: Opus, Sonnet, Haiku | | 86.4 | Claude 3 Opus | 2024-03-04 |
| Stable-DiffCoder: Pushing the Frontier of Code Diffusion Large Language Model | ✓ Link | 85.7 | Stable-DiffCoder-8B-Instruct | 2026-01-22 |
| Nemotron-Labs-Diffusion: A Tri-Mode Language Model Unifying Autoregressive, Diffusion, and Self-Speculation Decoding | | 85.45 | Nemotron-Labs-Diffusion-14B (Quad. SS) | 2026-07-07 |
| Latent Reasoning with Normalizing Flows | | 85.4 | NF-CoT (Unified) + RL | 2026-06-04 |
| Youtu-LLM: Unlocking the Native Agentic Potential for Lightweight Large Language Models | ✓ Link | 85 | Youtu-LLM 2B | 2025-12-31 |
| Planning-Driven Programming: A Large Language Model Programming Workflow | ✓ Link | 84.8 | LPW (GPT-4o) | 2024-11-21 |
| Motif 3: Technical Report | ✓ Link | 84.6 | Motif 3 Base (3-shot) | 2026-08-10 |
| RAAS: LLM Agentic System Architecture Search with GRPO | ✓ Link | 84.18 | RAAS | 2026-06-03 |
| Falcon-H1: A Family of Hybrid-Head Language Models Redefining Efficiency and Performance | ✓ Link | 83.86 | Falcon-H1-34B-Instruct | 2025-07-30 |
| SOEN-101: Code Generation by Emulating Software Process Models Using Large Language Model Agents | | 83.8±0.6 | GPT-3.5 Turbo + FlowGenScrum + Test | 2024-03-23 |
| Atlas: Orchestrating Heterogeneous Models and Tools for Multi-Domain Complex Reasoning | | 83.6 | Atlas (cluster) | 2026-01-07 |
| AFlow: Automating Agentic Workflow Generation | ✓ Link | 83.4 | AFlow(GPT-4o-mini) | 2024-10-14 |
| How Does Naming Affect LLMs on Code Analysis Tasks? | | 83.2 | GPT-3.5 Turbo (ChatGPT) | 2023-07-24 |
| Execution Guided Line-by-Line Code Generation | ✓ Link | 83.2 | EG-CFG (DeepSeek Coder 1.3b Instruct) | 2025-06-12 |
| MapCoder: Multi-Agent Code Generation for Competitive Problem Solving | ✓ Link | 83.1 | MapCoder (GPT-4) | 2024-05-18 |
| Think Anywhere in Code Generation | ✓ Link | 82.9 | Think-Anywhere (Qwen2.5-Coder-7B-Instruct) | 2026-03-31 |
| Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models | ✓ Link | 82.3 | o1-mini + Language Agent Tree Search (Hamming.ai) | 2023-10-06 |
| How Does Naming Affect LLMs on Code Analysis Tasks? | | 82 | GPT-4 (Bing Chat) | 2023-07-24 |
| ReflexiCoder: Teaching Large Language Models to Self-Reflect on Generated Code and Self-Correct It via Reinforcement Learning | ✓ Link | 81.8 | ReflexiCoder-8B (Single) | 2026-03-06 |
| Qwen3 Technical Report | ✓ Link | 81.4 | Qwen3-235B-A22B (Base) | 2025-05-14 |
| Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models | ✓ Link | 81.1 | GPT-3.5 Turbo + Language Agent Tree Search | 2023-10-06 |
| LLaDA MoE v2: Scaling Mixture-of-Experts Diffusion Language Models | | 81.03 | LLaDA MoE v2 Instruct | 2026-08-04 |
| BatCoder: Self-Supervised Bidirectional Code-Documentation Learning via Back-Translation | | 81 | BatCoder (7B) | 2026-01-30 |
| Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone | ✓ Link | 80.8 | Phi-3.5-MoE (3-shot) | 2024-04-22 |
| From Code to Correctness: Closing the Last Mile of Code Generation with Hierarchical Debugging | ✓ Link | 80.8 | MGDebugger (CodeQwen1.5) | 2024-10-02 |
| The Claude 3 Model Family: Opus, Sonnet, Haiku | | 80.4 | Claude 3 Haiku | 2024-03-04 |
| Teaching Large Language Models to Self-Debug | ✓ Link | 80.2 | GPT-4 (Self-Debugging with unit tests + trace) | 2023-04-11 |
| Qwen2 Technical Report | ✓ Link | 80.2 | Qwen2-72B-Instruct | 2024-07-15 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 80 | GPT-4 (few-shot) | 2024-01-25 |
| FlowEvo: Self-Evolving Agents through the Co-Evolution of Workflows and Executable Skills | ✓ Link | 79.6 | FlowEvo | 2026-08-20 |
| The Claude 3 Model Family: Opus, Sonnet, Haiku | | 79.4 | Claude 3 Sonnet | 2024-03-04 |
| Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters | ✓ Link | 79.4 | Step-3.5-Flash Base | 2026-02-11 |
| Safin-1: Safety from Within through Memory-Native State Evolution | ✓ Link | 78.6 | Safin-1-35B-A3B-SFT | 2026-08-31 |
| Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models | ✓ Link | 77.8 | Nemotron-H-56B-Base (3-shot) | 2025-04-04 |
| Locally Confident, Globally Stuck: The Quality-Exploration Dilemma in Diffusion Language Models | | 77.6 | IMH (WeDLM-8B) | 2026-04-01 |
| LoopRPT: Reinforcement Pre-Training for Looped Language Models | | 77.24 | Ouro (2.6B) LoopRPT | 2026-03-20 |
| Qwen2 Technical Report | ✓ Link | 76.9 | Qwen2-72B | 2024-07-15 |
| InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models | ✓ Link | 76.7 | InternVL3-78B | 2025-04-14 |
| MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction | ✓ Link | 76.7 | MiniCPM-o 4.5-Instruct | 2026-04-30 |
| Geometry Conflict: Explaining and Controlling Forgetting in LLM Continual Post-Training | ✓ Link | 76.7 | Qwen3-14B with GCWM | 2026-05-10 |
| How Does Naming Affect LLMs on Code Analysis Tasks? | | 76.2 | Bard (PaLM 2/chat-bison-001) | 2023-07-24 |
| Nemotron-4 340B Technical Report | ✓ Link | 75.4 | Nemotron-4-340B-Instruct | 2024-06-17 |
| Hybrid Policy Distillation for LLMs | ✓ Link | 75.4 | HPD (Qwen2.5-Coder, 1.5B) | 2026-04-22 |
| Understanding Evolution Strategies for LLM Reasoning: Broader Reasoning Coverage than GRPO | | 75.2 | Qwen2.5-7B-Instruct (ES to GRPO) | 2026-08-27 |
| UniSD: Towards a Unified Self-Distillation Framework for Large Language Models | ✓ Link | 74.7 | UniSD* | 2026-05-07 |
| Gemma 3 Technical Report | ✓ Link | 74.4 | Gemma 3 (27B, IT) (3-shot) | 2025-03-25 |
| Attention Residuals | ✓ Link | 73.9 | Kimi Linear (48B) with Block AttnRes | 2026-03-16 |
| Qwen2.5-Omni Technical Report | ✓ Link | 73.2 | Qwen2.5-Omni-7B | 2025-03-26 |
| Teaching Large Language Models to Self-Debug | ✓ Link | 72.8 | GPT-3.5 Turbo (Self-Debugging with unit tests + trace) | 2023-04-11 |
| OptiMer: Optimal Distribution Vector Merging Is Better than Data Mixing for Continual Pre-Training | ✓ Link | 72.6 | OptiMer | 2026-03-30 |
| ACES: Who Tests the Tests? Leave-One-Out AUC Consistency for Code Generation | | 72.37 | ACES-O | 2026-04-05 |
| DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model | ✓ Link | 72 | DeepSeek-V2 Chat (RL) (3-shot) | 2024-05-07 |
| Qwen2 Technical Report | ✓ Link | 71.9 | Qwen2-57B-A14B | 2024-07-15 |
| Group Sequence Policy Optimization | | 71.9 | AliceAI-T5-35B-A0.6B-Base | 2025-07-24 |
| Ministral 3 | | 71.6 | Ministral 3 (14B) | 2026-01-13 |
| How Does Naming Affect LLMs on Code Analysis Tasks? | | 71.4 | Claude | 2023-07-24 |
| Qwen2 Technical Report | ✓ Link | 70.9 | Qwen2-57B-A14B-Instruct | 2024-07-15 |
| Teaching Large Language Models to Self-Debug | ✓ Link | 70.8 | code-davinci-002 175B (Self-Debugging with unit tests + trace) | 2023-04-11 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 70.8 | GPT-3.5 Turbo (few-shot) | 2024-01-25 |
| Fast and Accurate Causal Parallel Decoding using Jacobi Forcing | ✓ Link | 70.4 | Jacobi Forcing Model (MR) | 2025-12-16 |
| A Sovereign, Open-Source Foundation Model for German and English | ✓ Link | 70.2 | Soofi S 30B-A3B Base | 2026-07-10 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 70 | DeepSeek-Coder-Instruct 33B (few-shot) | 2024-01-25 |
| INTERVENOR: Prompting the Coding Ability of Large Language Models with the Interactive Chain of Repair | ✓ Link | 69.8 | GPT-3.5 Turbo + INTERVENOR | 2023-11-16 |
| FineRMoE: Dimension Expansion for Finer-Grained Expert with Its Upcycling Approach | ✓ Link | 69.4 | FineRMoE (Qwen2.5-7B) | 2026-03-09 |
| LEVER: Learning to Verify Language-to-Code Generation with Execution | ✓ Link | 68.9 | code-davinci-002 175B + LEVER | 2023-02-16 |
| S2D2: Fast Decoding for Diffusion LLMs via Training-Free Self-Speculation | ✓ Link | 68.8 | S2D2 (LLaDA2.1-Mini) | 2026-03-26 |
| ReFusion: A Diffusion Large Language Model with Parallel Autoregressive Decoding | ✓ Link | 68.2 | ReFusion | 2025-12-15 |
| CodeT: Code Generation with Generated Tests | ✓ Link | 67.7 | code-davinci-002 175B + CodeT | 2022-07-21 |
| Teaching Large Language Models to Self-Debug | ✓ Link | 67.6 | GPT-3.5 Turbo (3-shot) | 2023-04-11 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 66.9 | code-davinci-002 175B + Reviewer | 2022-11-29 |
| T5Gemma 2: Seeing, Reading, and Understanding Longer | | 66.8 | T5Gemma 2 4B-4B (post-trained) | 2025-12-16 |
| NeuPAT: Neuron-aware Plasticity Allocation Tuning for Language-Preserving MLLMs | | 66.8 | NeuPAT (Qwen2.5-14B-Instruct) (extra data) | 2026-08-08 |
| CharacterFlywheel: Scaling Iterative Improvement of Engaging and Steerable LLMs in Production | | 66.6 | CharacterFlywheel V7 (Llama 3.1 70B) | 2026-03-02 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 66.4 | code-davinci-002 175B + Coder-Reviewer | 2022-11-29 |
| Grammar-Constrained Decoding Can Jailbreak LLMs into Generating Malicious Code | ✓ Link | 66.4 | CodeShield (Qwen2.5-Coder-7B) | 2026-06-10 |
| StarCoder 2 and The Stack v2: The Next Generation | ✓ Link | 66.2 | StarCoder2-15B | 2024-02-29 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 66 | DeepSeek-Coder-Base 33B (few-shot) | 2024-01-25 |
| Co-RL: Unsupervised Reasoning Emerges from Diverse Cohort in Multi-agent RL | ✓ Link | 65.6 | Co-RL (Qwen2.5-7B, Different family) | 2026-08-19 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 65.5 | Code Llama - Python 70B (3-shot) | 2023-08-24 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 65.4 | DeepSeek-Coder-Instruct 6.7B (few-shot) | 2024-01-25 |
| Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models | ✓ Link | 65.4 | Nemotron-H-8B-Base (3-shot) | 2025-04-04 |
| Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs | ✓ Link | 65.3 | Phi-4-Mini (3-shot) | 2025-02-24 |
| On the Off-Policy Teacher in On-Policy Distillation | | 65.2 | SCOUT | 2026-09-29 |
| Encoder-Decoder Gemma: Improving the Quality-Efficiency Trade-Off via Adaptation | ✓ Link | 64.8 | T5Gemma 9B-9B PrefixLM IT | 2025-04-08 |
| Self-Execution Simulation Improves Coding Models | | 63.2 | Qwen2.5-7B + NLEX | 2026-03-11 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 63 | code-davinci-002 175B + MBR-Exec | 2022-11-29 |
| Gemma 2: Improving Open Language Models at a Practical Size | ✓ Link | 62.6 | Gemma 2 27B (3-shot) | 2024-07-31 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 62.4 | Code Llama 70B (3-shot) | 2023-08-24 |
| Mellum2 Technical Report | ✓ Link | 62.4 | Mellum 2 (base) | 2026-05-29 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 62.2 | Code Llama - Instruct 70B (3-shot) | 2023-08-24 |
| CodeT: Code Generation with Generated Tests | ✓ Link | 61.9 | code-davinci-001 175B + CodeT | 2022-07-21 |
| Teaching Large Language Models to Self-Debug | ✓ Link | 61.4 | code-davinci-002 175B (3-shot) | 2023-04-11 |
| DeepSeek LLM: Scaling Open-Source Language Models with Longtermism | ✓ Link | 61.4 | DeepSeek LLM 67B Chat | 2024-01-05 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 61.2 | Unnatural Code Llama 34B (3-shot) | 2023-08-24 |
| Mixtral of Experts | ✓ Link | 60.7 | Mixtral 8x7B (3-shot) | 2024-01-08 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 60.6 | DeepSeek-Coder-Base 6.7B (few-shot) | 2024-01-25 |
| CERA-MoA: Co-Evolving Routing Mechanisms with Continually Learning LLM Agents | ✓ Link | 60.3 | CERA-MoA (Qwen3-4B) | 2026-09-16 |
| dOPSD: On-Policy Self-Distillation for Diffusion Language Models | ✓ Link | 58.49 | Dream-7B-Instruct (dOPSD) | 2026-07-05 |
| Natural Language to Code Translation with Execution | ✓ Link | 58.2 | code-davinci-001 175B + MBR-Exec | 2022-04-25 |
| Improved Large Language Diffusion Models | ✓ Link | 58 | iLLaDA 8B Instruct | 2026-06-24 |
| Improved Large Language Diffusion Models | ✓ Link | 57.8 | iLLaDA 8B Base | 2026-06-24 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 57 | Code Llama - Instruct 34B (3-shot) | 2023-08-24 |
| Every Token Counts: Generalizing 16M Ultra-Long Context in Large Language Models | ✓ Link | 57 | HSA-UltraLong-Inst | 2025-11-28 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 56.2 | Code Llama - Python 34B (3-shot) | 2023-08-24 |
| Encoder-Decoder Gemma: Improving the Quality-Efficiency Trade-Off via Adaptation | ✓ Link | 55.6 | T5Gemma 9B-9B PrefixLM | 2025-04-08 |
| Textbooks Are All You Need | ✓ Link | 55.5 | Phi-1 | 2023-06-20 |
| CodeT: Code Generation with Generated Tests | ✓ Link | 55.4 | code-cushman-001 12B (CodeT) | 2022-07-21 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 55 | Code Llama 34B (3-shot) | 2023-08-24 |
| Teaching Large Language Models to Self-Debug | ✓ Link | 53.2 | StarCoder 15.5B (Self-Debugging with unit tests + trace) | 2023-04-11 |
| StarCoder: may the source be with you! | ✓ Link | 52.7 | StarCoder 15.5B | 2023-05-09 |
| The Flexibility Trap: Why Arbitrary Order Limits Reasoning Potential in Diffusion Language Models | ✓ Link | 52.4 | JustGRPO (LLaDA 8B Instruct) | 2026-01-21 |
| MobileMoE: Scaling On-Device Mixture of Experts | | 52.4 | MobileMoE-L (922M) | 2026-05-26 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 52.2 | GPT-3.5 Turbo | 2023-08-24 |
| WizardCoder: Empowering Code Large Language Models with Evol-Instruct | ✓ Link | 51.8 | WizardCoder 15B | 2023-06-14 |
| Introducing LFM-7B: Setting New Standards for Efficient Language Models | | 51.6 | LFM-7B | 2025-01-20 |
| PaLM 2 Technical Report | ✓ Link | 50 | PaLM 2-S* (few-shot) | 2023-05-17 |
| CodeT: Code Generation with Generated Tests | ✓ Link | 49.5 | CodeGen-Mono 16B + CodeT | 2022-07-21 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 49.4 | Code Llama - Instruct 13B (3-shot) | 2023-08-24 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 49.4 | DeepSeek-Coder-Instruct 1.3B (few-shot) | 2024-01-25 |
| StarCoder: may the source be with you! | ✓ Link | 49 | StarCoderBase 15.5B | 2023-05-09 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 49 | Code Llama - Python 13B (3-shot) | 2023-08-24 |
| daVinci-LLM:Towards the Science of Pretraining | ✓ Link | 49 | daVinci-LLM-3B | 2026-03-28 |
| Parameter-Efficient Sparsity Crafting from Dense to Mixture-of-Experts for Instruction Tuning on General Tasks | ✓ Link | 48.6 | Qwen2idae-16x14B (4-shot) | 2024-01-05 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 48.3 | code-cushman-001 12B + MBR-Exec | 2022-11-29 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 47.6 | Code Llama - Python 7B (3-shot) | 2023-08-24 |
| Mistral 7B | ✓ Link | 47.5 | Mistral 7B (3-shot) | 2023-10-10 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 47.3 | CodeGen 16B + MBR-Exec | 2022-11-29 |
| Teaching Large Language Models to Self-Debug | ✓ Link | 47.2 | StarCoder 15.5B (3-shot) | 2023-04-11 |
| PaLM: Scaling Language Modeling with Pathways | ✓ Link | 47 | PaLM Coder 540B | 2022-04-05 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 47 | Code Llama 13B (3-shot) | 2023-08-24 |
| Apertus: Democratizing Open and Compliant LLMs for Global Language Environments | ✓ Link | 47 | Apertus-70B-Instruct | 2025-09-17 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 46.2 | CodeGen 16B + Coder-Reviewer | 2022-11-29 |
| DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence | ✓ Link | 46.2 | DeepSeek-Coder-Base 1.3B (few-shot) | 2024-01-25 |
| EpiCaR: Knowing What You Don't Know Matters for Better Reasoning in LLMs | | 45.91 | EpiCaR (Qwen-3, 8B) | 2026-01-11 |
| dLLM: Simple Diffusion Language Modeling | ✓ Link | 45.53 | Dream-Instruct (reasoning SFT) | 2026-02-26 |
| INTERVENOR: Prompting the Coding Ability of Large Language Models with the Interactive Chain of Repair | ✓ Link | 45.4 | GPT-3.5 Turbo (few-shot) | 2023-11-16 |
| Llama 2: Open Foundation and Fine-Tuned Chat Models | ✓ Link | 45 | Llama 2 70B (zero-shot) | 2023-07-18 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 44.4 | Code Llama - Instruct 7B (3-shot) | 2023-08-24 |
| Gemma: Open Models Based on Gemini Research and Technology | ✓ Link | 44.4 | Gemma 7B (3-shot) | 2024-03-13 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 44.1 | CodeGen 16B + Reviewer | 2022-11-29 |
| Textbooks Are All You Need II: phi-1.5 technical report | ✓ Link | 43.5 | phi-1.5-web 1.3B | 2023-09-11 |
| Branch-Train-MiX: Mixing Expert LLMs into a Mixture-of-Experts LLM | ✓ Link | 42.6 | Branch-Train-Merge 4x7B (top-2) | 2024-03-12 |
| Stop the Flip-Flop: Context-Preserving Verification for Fast Revocable Diffusion Decoding | | 42.4 | COVER (LLaDA-Base-8B) | 2026-02-05 |
| RecurrentGemma: Moving Past Transformers for Efficient Open Language Models | ✓ Link | 42 | RecurrentGemma-9B (3-shot) | 2024-04-11 |
| Byte Latent Transformer: Patches Scale Better Than Tokens | ✓ Link | 41.8 | BLT-Entropy 8B (3-shot) | 2024-12-13 |
| Code Llama: Open Foundation Models for Code | ✓ Link | 41.4 | Code Llama 7B (3-shot) | 2023-08-24 |
| Parameter-Efficient Sparsity Crafting from Dense to Mixture-of-Experts for Instruction Tuning on General Tasks | ✓ Link | 41.4 | Camelidae-8×34B (4-shot) | 2024-01-05 |
| Dystruct: Dynamically Structured Diffusion Language Model Decoding via Bayesian Inference | | 41.4 | DyStruct (LLaDA-8B-Base) | 2026-05-10 |
| Nemotron-4 15B Technical Report | ✓ Link | 40.6 | Nemotron-4 15B (3-shot) | 2024-02-26 |
| INTERVENOR: Prompting the Coding Ability of Large Language Models with the Interactive Chain of Repair | ✓ Link | 39.8 | GPT-3.5 Turbo (0-shot) | 2023-11-16 |
| Branch-Train-MiX: Mixing Expert LLMs into a Mixture-of-Experts LLM | ✓ Link | 39.4 | Branch-Train-MiX 4x7B (sampling top-2 experts) | 2024-03-12 |
| Turning the TIDE: Cross-Architecture Distillation for Diffusion Large Language Models | ✓ Link | 38.6 | Qwen3-0.6B-BD3LM (Tide-Shared) | 2026-04-29 |
| Prism: Efficient Test-Time Scaling via Hierarchical Search and Self-Verification for Discrete Diffusion Language Models | ✓ Link | 38.2 | Prism (LLaDA-8B-Instruct, K=8) | 2026-02-02 |
| LLaMA: Open and Efficient Foundation Language Models | ✓ Link | 37.7 | LLaMA 65B (0-shot) | 2023-02-27 |
| PaLM: Scaling Language Modeling with Pathways | ✓ Link | 36.8 | PaLM 540B | 2022-04-05 |
| Token-Level LLM Collaboration via FusionRoute | | 36 | FusionRoute (Llama-3.1-8B) | 2026-01-08 |
| DeepSeek-VL: Towards Real-World Vision-Language Understanding | ✓ Link | 35.2 | DeepSeek-VL 7B Chat | 2024-03-08 |
| Compact Language Models via Pruning and Knowledge Distillation | ✓ Link | 35.2 | Minitron 8B | 2024-07-19 |
| SantaCoder: don't reach for the stars! | ✓ Link | 35 | SantaCoder | 2023-01-09 |
| StarCoder: may the source be with you! | ✓ Link | 35 | SantaCoder 1.1B | 2023-05-09 |
| CodeT: Code Generation with Generated Tests | ✓ Link | 34.4 | InCoder 6.7B + CodeT | 2022-07-21 |
| Power Scheduler: A Batch Size and Token Number Agnostic Learning Rate Scheduler | ✓ Link | 33.6 | PowerLM-3B | 2024-08-23 |
| Llama 2: Open Foundation and Fine-Tuned Chat Models | ✓ Link | 33 | Llama 2 34B (0-shot) | 2023-07-18 |
| Llama 2: Open Foundation and Fine-Tuned Chat Models | ✓ Link | 30.6 | Llama 2 13B (0-shot) | 2023-07-18 |
| LLaMA: Open and Efficient Foundation Language Models | ✓ Link | 30.2 | LLaMA 33B (0-shot) | 2023-02-27 |
| RecurrentGemma: Moving Past Transformers for Efficient Open Language Models | ✓ Link | 28.8 | RecurrentGemma-2B (3-shot) | 2024-04-11 |
| Compact Language Models via Pruning and Knowledge Distillation | ✓ Link | 28.2 | Minitron 4B | 2024-07-19 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 26.7 | InCoder 6.7B + MBR-Exec | 2022-11-29 |
| Sumi: Open Uniform Diffusion Language Model from Scratch | ✓ Link | 26.6 | Sumi-7B (3-shot) | 2026-06-17 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 26.1 | InCoder 6.7B + Coder-Reviewer | 2022-11-29 |
| Coder Reviewer Reranking for Code Generation | ✓ Link | 24.4 | InCoder 6.7B + Reviewer | 2022-11-29 |
| CodeGeeX: A Pre-Trained Model for Code Generation with Multilingual Benchmarking on HumanEval-X | ✓ Link | 24.4 | CodeGeeX-13B | 2023-03-30 |
| LLaMA: Open and Efficient Foundation Language Models | ✓ Link | 22 | LLaMA 13B (0-shot) | 2023-02-27 |
| Llama 2: Open Foundation and Fine-Tuned Chat Models | ✓ Link | 20.8 | Llama 2 7B (0-shot) | 2023-07-18 |
| Copy-Paste Language Models | ✓ Link | 20.8 | Copy-Paste Language Model (1.7B, SW-256) | 2026-10-13 |
| InCoder: A Generative Model for Code Infilling and Synthesis | ✓ Link | 19.4 | InCoder 6.7B (0-shot) | 2022-04-12 |
| LLaMA: Open and Efficient Foundation Language Models | ✓ Link | 17.7 | LLaMA 7B (0-shot) | 2023-02-27 |
| Scaling Inherently Interpretable Language Models | ✓ Link | 11 | Steerling-8B | 2026-08-06 |