AI Paper Atlas

Research map / Benchmarks / MRPC

MRPC leaderboard

63 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on MRPC. Best accuracy: 93.32 (KNAS, ICML 2021).

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.

MRPC · accuracy

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

#accuracy ↑MethodPaperTopic
193.32KNAS
RoBERTA-large
KNAS: Green Neural Architecture Search
ICML 2021 · Table 2
Neural architecture search
290.69M-SMoE
switch-base-32
Merge, Then Compress: Demystify Efficient SMoE with Hints from Its Routing Policy
ICLR 2024 · Table 1
Mixture of experts
390.1KNIFE
BERT
A Differential Entropy Estimator for Training Neural Networks
ICML 2022 · Table: Fine-tuning on GLUE
Variational and generative sampling
490rdFFT p=512
RoBERTa-large
Memory-Efficient Training with In-Place FFT Implementation
NeurIPS 2025 · Model-level evaluation table
RNNs and normalization
589.95LoRAM
DeBERTa-v3-base
The Primacy of Magnitude in Low-Rank Adaptation
NeurIPS 2025 · GLUE results table
Pruning and efficient tuning
689.59Flat-LoRA
T5-Base
Flat-LoRA: Low-Rank Adaptation over a Flat Loss Landscape
ICML 2025 · Table: Comparison with other L
Pruning and efficient tuning
789.23VIBERT Base
BERT Base
Variational Information Bottleneck for Effective Low-Resource Fine-Tuning
ICLR 2021 · GLUE results table
Recognition and long-tail
889.22LSTSD
BERT
Long Short-Term Sample Distillation
AAAI 2020 · GLUE results table
Knowledge distillation
989.2ALBERT-base+BABN
ALBERT-base
Bayesian Attention Belief Networks
ICML 2021 · Table 1
Uncertainty and Gaussian processes
1089.1Opt-SAM
Bert-Large
Improving Sharpness-Aware Minimization by Lookahead
ICML 2024 · Table 6
Implicit bias and generalization
37 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

MRPC · F1

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

#F1 ↑MethodPaperTopic
193.9BOMF
RoBERTa-base
Model Fusion through Bayesian Optimization in Language Model Fine-Tuning
NeurIPS 2024 · Table 1
Bayesian optimization
293.5SUMO (SVD, rank=4)
RoBERTa-Base
SUMO: Subspace-Aware Moment-Orthogonalization for Accelerating Memory-Efficient LLM Training
NeurIPS 2025 · GLUE results table
Optimizers
393RoBERTa Large + AT + CTR
RoBERTa Large
Improved Text Classification via Contrastive Adversarial Training
AAAI 2022 · Table 2
Certified and empirical robustness
492.6AdaRankGrad (initial rank=4)
RoBERTa-Base
AdaRankGrad: Adaptive Gradient Rank and Moments for Memory-Efficient LLMs Training and Fine-Tuning
ICLR 2025 · Table: GLUE benchmark
Pruning and efficient tuning
592.2DynaKD
6-layer BERT-style model
How to Trade Off the Quantity and Capacity of Teacher Ensemble: Learning Categorical Distribution to Stochastically Employ a Teacher for Distillation
AAAI 2024 · Table 5
Knowledge distillation
692.1PortLLM
Mistral-7B
PortLLM: Personalizing Evolving Large Language Models with Training-Free and Portable Model Patches
ICLR 2025 · Table: Zero-shot performance c
Instruction and fine-tuning
791.4Re-init
BERT-large
Revisiting Few-sample BERT Fine-tuning
ICLR 2021 · Table: Standard vs. Re-init an
Recognition and long-tail
891.2Few-shot, 3 partitions
BERT-base
Few-shot Task-agnostic Neural Architecture Search for Distilling Large Language Models
NeurIPS 2022 · Table 6
Neural architecture search
990.3FedAbA
RoBERTa-large
Align Before Aggregation: Basis-Consistent Federated LoRA under Heterogeneous Ranks
NeurIPS 2026 · Table 1
Personalized and heterogeneous FL
1089.13Mr.BiQ W2A8
BERT
Mr.BiQ: Post-Training Non-Uniform Quantization Based on Minimizing the Reconstruction Error
CVPR 2022 · Table 6
Quantization
15 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.