AI Paper Atlas

Research map / Benchmarks / GLUE average

GLUE average leaderboard

13 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on GLUE average. Best average score: 91.3 (DeBERTa XXL LoRA, ICLR 2022).

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.

GLUE average · average score

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

#average score ↑MethodPaperTopic
191.3DeBERTa XXL LoRA
DeBERTa XXL
LoRA: Low-Rank Adaptation of Large Language Models
ICLR 2022 · GLUE results table
Pruning and efficient tuning
290.53LoPA
RoBERTa 355M
Prompt Tuning Strikes Back: Customizing Foundation Models with Low-Rank Prompt Adaptation
NeurIPS 2024 · GLUE results table
Instruction and fine-tuning
389.8LoRA+WTA-CRS@0.3
T5-3B
Winner-Take-All Column Row Sampling for Memory Efficient Adaptation of Language Model
NeurIPS 2023 · Table 1
Optimizers
489.1MoLEx
RoBERTa-large
MoLEx: Mixture of Layer Experts for Fine-tuning with Sparse Upcycling
ICLR 2025 · Table 1
Pruning and efficient tuning
589.1SVFT Random d=2
DeBERTaV3-base
SVFT: Parameter-Efficient Fine-Tuning with Singular Vectors
NeurIPS 2024 · Table: GLUE
Pruning and efficient tuning
687.17S3Delta-M (1.39% params)
T5-large
Sparse Structure Search for Delta Tuning
NeurIPS 2022 · Table 2
Pruning and efficient tuning
785.3D2NWG
RoBERTa-Base
Diffusion-based Neural Network Weights Generation
ICLR 2025 · Task Conditioned LoRA paramete
Fast diffusion sampling
883.4AD-KD + OPDF
BERT-base-uncased
Over-parameterized Student Model via Tensor Decomposition Boosted Knowledge Distillation
NeurIPS 2024 · Table 1
RNNs and normalization
982.3CubeBERT128
BERT
What Dense Graph Do You Need for Self-Attention?
ICML 2022 · Table: Performances on GLUE te
Attention and transformers
1077.6SpikeLM + AdaS
SpikeLM
AdaS: Adaptive Gradient Descent for Spiking Transformers
ICML 2026 · Table 1
Spiking networks
3 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.