AI Paper Atlas

Research map / Benchmarks / AMBER

AMBER leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on AMBER. Best CHAIR: 2.1 (TARS (Mask), ECCV 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.

AMBER · CHAIR

10 papers, each with its best reported number for its own method; lower is better.

#CHAIR ↓MethodPaperTopic
12.1TARS (Mask)
LLaVA-v1.5-13B
TARS: MinMax Token-Adaptive Preference Strategy for Hallucination Reduction in MLLMs
ECCV 2026 · Table 1
Multimodal LLMs
22.2OPA-DPO
LLaVA-Instruct-1.5-7B
Mitigating Hallucinations in Large Vision-Language Models via DPO: On-Policy Data Hold the Key
CVPR 2025 · Benchmark Evaluation Results t
Multimodal LLMs
32.4IRIS-R2
LLaVA-1.5 7B
IRIS: Implicit Reward-Guided Internal Sifting for Mitigating Multimodal Hallucination
ICML 2026 · Table 1
Multimodal LLMs
42.9HalMask
LLaVA-1.5-7B
Learning to Mask: Cross-Modal Noise Modulation for Hallucination Mitigation in Multi-modal Large Language Models
ECCV 2026 · Table 4
Multimodal LLMs
52.9SENTINEL
LLaVA-v1.5-7B
Mitigating Object Hallucinations via Sentence-Level Early Intervention
ICCV 2025 · Table 1
Multimodal LLMs
64.6CRG
LLaVA-1.5-7B
Mitigating Hallucinations in Large Vision-Language Models via Causal Route Gating
ICML 2026 · Table 3
Multimodal LLMs
74.8Octopus
LLaVA-v1.5-7B
Octopus: Alleviating Hallucination via Dynamic Contrastive Decoding
CVPR 2025 · Table 1
Multimodal LLMs
86.6HALVA-Med
LLaVA-v1.5-7B
Mitigating Object Hallucination in MLLMs via Data-augmented Phrase-level Alignment
ICLR 2025 · Table 2
Multimodal LLMs
97.6MGAP (Ours)
LLaVA v1.5 7B
Mitigating Manifold Departure: Uncertainty-Aware Subspace Rectification for Trustworthy MLLM Decoding
ICML 2026 · Table 5
Multimodal LLMs
1017.8KVSmooth
MiniGPT-4 7B
KVSmooth: Mitigating Hallucination in Multi-modal Large Language Models through Key-Value Smoothing
CVPR 2026 · Table 3
Multimodal LLMs

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