AI Paper Atlas

Research map / Benchmarks / ScienceQA-IMG

ScienceQA-IMG leaderboard

36 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on ScienceQA-IMG. Best accuracy: 97.3 (T-DRS, AAAI 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.

ScienceQA-IMG · accuracy

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

#accuracy ↑MethodPaperTopic
197.3T-DRS
InterVL2-8B
Remember Me: Bridging the Long-Range Gap in LVLMs with Three-Step Inference-Only Decay Resilience Strategies
AAAI 2026 · Table 1
Multimodal LLMs
293.5MUSE-VL
Qwen2.5-7B
MUSE-VL: Modeling Unified VLM through Semantic Discrete Encoding
ICCV 2025 · Table 2
Multimodal LLMs
391.7ETT
Qwen2.5-1.5B
End-to-End Vision Tokenizer Tuning
NeurIPS 2025 · Multimodal understanding table
Human and embodied modeling
490.5Prophet++ (mPLUG)Prophet: Prompting Large Language Models With Complementary Answer Heuristics for Knowledge-Based Visual Question Answering
TPAMI 2025 · Table 5a
Visual question answering
584.18SliME-8B
Llama3-8B
Beyond LLaVA-HD: Diving Into High-Resolution Multimodal Large Language Models
TPAMI 2026 · Table 2
Multimodal LLMs
680.5REVERIE-1.5-7b-lora
LLaVA-1.5-7b-lora
Reflective Instruction Tuning: Mitigating Hallucinations in Large Vision-Language Models
ECCV 2024 · Table 2
Multimodal LLMs
775.8GPT-4V + CoCT (CoBRa)
GPT-4V
Learning Chain of Counterfactual Thought for Bias-Robust Vision-Language Reasoning
ECCV 2024 · Table 2
Multimodal LLMs
873.7LLaVA-1.5+HyperET
Vicuna-13B
HyperET: Efficient Training in Hyperbolic Space for Multi-modal Large Language Models
NeurIPS 2025 · Table: Comparison with SoTA pr
Multimodal LLMs
973.7VILA-13B
Llama-2-13B
VILA: On Pre-training for Visual Language Models
CVPR 2024 · Table 5
CLIP and zero-shot
1070.9STGC
MoE-LLaVA, Phi-2-2.7B, 4Top2
Solving Token Gradient Conflict in Mixture-of-Experts for Large Vision-Language Model
ICLR 2025 · Table 1
Mixture of experts
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

ScienceQA-IMG · score

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

#score ↑MethodPaperTopic
196.9HaploVL-7B-Pro
Qwen2.5-7B
HaploVL: A Single-Transformer Baseline for Multi-Modal Understanding
ICML 2025 · Main results table
Language, safety and interpretability
296.2EVEv2.0 (Ours)
Qwen2.5-7B
EVEv2: Improved Baselines for Encoder-Free Vision-Language Models
ICCV 2025 · Table 2 (main comparison)
Multimodal LLMs
393.3SAIL
Mistral-7B-v0.1
The Scalability of Simplicity: Empirical Analysis of Vision-Language Learning with a Single Transformer
ICCV 2025 · Table 2
Cross-modal learning
478HAMSTER
VILA1.5-13B
HAMSTER: Hierarchical Action Models for Open-World Robot Manipulation
ICLR 2025 · Table: Core VQA and robustness
Robot action and navigation
573.1BASIC
Vicuna-13B
BASIC: Boosting Visual Alignment with Intrinsic Refined Embeddings in Multimodal Large Language Models
ICCV 2025 · Table 1
Multimodal LLMs
672.1LLaVA-S² (Ours)
LLaVA-S², Vicuna-7B
DenseFusion-1M: Merging Vision Experts for Comprehensive Multimodal Perception
NeurIPS 2024 · Main results table
Multimodal LLMs
770.9Mipha-3B
Phi-2 (2.7B)
A Comprehensive Overhaul of Multimodal Assistant with Small Language Models
AAAI 2025 · Table: Multi-modal evaluation
Multimodal LLMs
869.66Blink
LLaVA-1.5-7B
Blink: Dynamic Visual Token Resolution for Enhanced Multimodal Understanding
CVPR 2026 · Table 1
Multimodal LLMs
969.51ETA
LLaVA-1.5-7B
ETA: Evaluating Then Aligning Safety of Vision Language Models at Inference Time
ICLR 2025 · General performance table
Multimodal LLMs
1068MaTVLM Hybrid-Mamba-25%
TinyLLaVA-Phi-2-SigLIP-3.1B
MaTVLM: Hybrid Mamba-Transformer for Efficient Vision-Language Modeling
ICCV 2025 · Table 1
Attention and transformers
1 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.