AI Paper Atlas

Research map / Benchmarks / AlpacaEval

AlpacaEval leaderboard

20 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on AlpacaEval. Best win rate: 94.78 (Mistral-7B-ReMax, ICML 2024).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

AlpacaEval · win rate

20 papers, each with its best reported number for its own method; higher is better.

#win rate ↑MethodPaperTopic
194.78Mistral-7B-ReMax
Mistral-7B-Instruct-v0.2
ReMax: A Simple, Effective, and Efficient Reinforcement Learning Method for Aligning Large Language Models
ICML 2024 · Table: Performance against str
Alignment and preferences
289.5OpenChat-13B
Llama-2-13B
OpenChat: Advancing Open-source Language Models with Mixed-Quality Data
ICLR 2024 · Main Results table
Instruction and fine-tuning
388.32Dromedary-2 70B
LLaMA-2-70b
SALMON: Self-Alignment with Instructable Reward Models
ICLR 2024 · Appendix, AlpacaEval table
Alignment and preferences
486.3UltraLM-13B-PPO
UltraLM-13B
ULTRAFEEDBACK: Boosting Language Models with Scaled AI Feedback
ICML 2024 · Head-to-head comparison table
Alignment and preferences
573.01CoE-3E2A
LLaMA2-7B-Base
Fine-Tuning Language Models with Collaborative and Semantic Experts
AAAI 2025 · Table 2
Instruction and fine-tuning
672.5TIES-Merging, WizardLM-13B + WizardMath-13B + llama-2-13b-code-alpaca
Llama-2-13b
Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunch
ICML 2024 · Table: Performance of merging
Instruction and fine-tuning
763.5DROOn Prompt-Driven Safeguarding for Large Language Models
ICML 2024 · Table: Evaluation results (def
Language, safety and interpretability
863RobustKV
Llama2-Chat-7B
RobustKV: Defending Large Language Models against Jailbreak Attacks via KV Eviction
ICLR 2025 · Short-text utility table
Attacks on text and graphs
955.65PT + LoRA SFT + top-k + Glob + Conv
LLaDA-8B-Base
Fast and Fluent Diffusion Language Models via Convolutional Decoding and Rejective Fine-tuning
NeurIPS 2025 · Comparison between large (8B)
Language, safety and interpretability
1055OnlineRubrics-pi_old
Qwen2.5-7B-Instruct
Online Rubrics Elicitation from Pairwise Comparisons
ICML 2026 · Table 2
Alignment and preferences
10 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.