AI Paper Atlas

Research map / Benchmarks / CB

CB leaderboard

12 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CB. Best accuracy: 96.43 (Dominant-layer MeZO, NeurIPS 2026).

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.

CB · accuracy

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

#accuracy ↑MethodPaperTopic
196.43Dominant-layer MeZO
Qwen3-8B
Dominant-Layer ZO: A Single Layer Dominates Zeroth-Order Fine-Tuning of LLMs
NeurIPS 2026 · Table 2
Instruction and fine-tuning
289.2AGZO
Qwen3-0.6B
AGZO: Activation-Guided Zeroth-Order Optimization for LLM Fine-Tuning
ICML 2026 · Table 1
Pruning and efficient tuning
387.5AR1-ZO (r=64)
Qwen3-1.7B
AR1-ZO: Topology-Aware Rank-1 Zeroth-Order Queries for High-Rank LoRA Fine-Tuning
NeurIPS 2026 · Table 1 (Qwen3-1.7B)
Pruning and efficient tuning
486.2DeBERTa-large + co-training
DeBERTa-large
Co-training Improves Prompt-based Learning for Large Language Models
ICML 2022 · Few-shot learning results tabl
Zero-/few-shot and meta-learning
583.9GPT-J-6B AMA
GPT-J-6B
Ask Me Anything: A simple strategy for prompting language models
ICLR 2023 · Table 1
Language, safety and interpretability
676.79Bilevel-ZOFO
OPT-1.3B
Bilevel ZOFO: Efficient LLM Fine-Tuning and Meta-Training
NeurIPS 2025 · Design-choice ablation table
Instruction and fine-tuning
776ZO Fine-tuner
LLaMA-3.1-8B
Learning a Zeroth-Order Optimizer for Fine-Tuning LLMs
ICML 2026 · Table 1
Instruction and fine-tuning
875.71DeComFL (P=10)
OPT-1.3B
Achieving Dimension-Free Communication in Federated Learning via Zeroth-Order Optimization
ICLR 2025 · Test Accuracy and Communicatio
Communication-efficient FL
971.8DiZO LoRA
OPT-2.7B
Harmony in Divergence: Towards Fast, Accurate, and Memory-efficient Zeroth-order LLM Fine-tuning
NeurIPS 2025 · Table: OPT-2.7B results
Instruction and fine-tuning
1071.4LoHO-SGDinter
OPT-13B
Towards Efficient Low-Order Hybrid Optimizer for Language Model Fine-Tuning
AAAI 2025 · Table 2
Pruning and efficient tuning
2 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.