AI Paper Atlas

Research map / Benchmarks / HotpotQA

HotpotQA leaderboard

146 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on HotpotQA. Best EM: 88.58 (CF-RAG, ICLR 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.

HotpotQA · EM

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

#EM ↑MethodPaperTopic
188.58CF-RAG
Llama-3-8B-Instruct
Counterfactual Reasoning for Retrieval-Augmented Generation
ICLR 2026 · Table 1
Retrieval and knowledge
280.3PACE
Qwen3.5-9B
PACE: Two-Timescale Self-Evolution for Small Language Model Agents
NeurIPS 2026 · Table 1
LLM agents
379.1DRAFT-RL
Claude-3.5-Sonnet
DRAFT-RL: Multi-Agent Chain-of-Draft Reasoning for Reinforcement Learning-Enhanced LLMs
AAAI 2026 · Table 3
LLM agents
478.12FlowSteer
GPT-4o-mini backend
FlowSteer: Towards Agents Designing Agentic Workflows via Reinforced Progressive Canvas Editing
NeurIPS 2026 · Table 2
LLM agents
574AgentTTS
GPT-o3-mini search agent
AgentTTS: Large Language Model Agent for Test-time Compute-optimal Scaling Strategy in Complex Tasks
NeurIPS 2025 · Table: Comparison of search ti
Reasoning and chain of thought
673.3LazyAttention
Tulu3-Block-FT (Llama-3.1-8B)
LazyAttention: Efficient Retrieval-Augmented Generation with Deferred Positional Encoding
ICML 2026 · Table 2
Retrieval and knowledge
772.1KG-Translator (Llama-3.2-1B)
Llama-3.2-1B
From Retrieval to Translation: Translating Query into Graph-level Clues for Retrieval-Augmented Generation
ICML 2026 · Table 1
Retrieval and knowledge
871LATS (CoT + ReAct)
GPT-3.5
Language Agent Tree Search Unifies Reasoning, Acting, and Planning in Language Models
ICML 2024 · Table 3
LLM agents
968.7Ours (Predictive)
Llama-3.1-8B
Predictive Prefetching for Retrieval-Augmented Generation
ICML 2026 · Table 2
Retrieval and knowledge
1068INSES + Router
GLM-4
Beyond Explicit Edges: Robust Reasoning over Noisy and Sparse Knowledge Graphs
ICML 2026 · Table 2
Knowledge and temporal graphs
64 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

HotpotQA · F1

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

#F1 ↑MethodPaperTopic
190.5DRAFT-RL
Claude-3.5-Sonnet
DRAFT-RL: Multi-Agent Chain-of-Draft Reasoning for Reinforcement Learning-Enhanced LLMs
AAAI 2026 · Table 3
LLM agents
284.98FlowSteer
GPT-4o-mini backend
FlowSteer: Towards Agents Designing Agentic Workflows via Reinforced Progressive Canvas Editing
NeurIPS 2026 · Table 2
LLM agents
383.4SAIN SpanBERT
SpanBERT
Semantics-Aware Inferential Network for Natural Language Understanding
AAAI 2021 · Table 2
Language, safety and interpretability
483.22InfoRE
GPT-4
Information Re-Organization Improves Reasoning in Large Language Models
NeurIPS 2024 · Table 2
Retrieval and knowledge
581.07KG-Translator (Qwen-2.5-1.5B)
Qwen-2.5-1.5B
From Retrieval to Translation: Translating Query into Graph-level Clues for Retrieval-Augmented Generation
ICML 2026 · Table 1
Retrieval and knowledge
678.8XBridge
Llama-3.1-8B -> Qwen2.5-7B
XBRIDGE: Entity-Grounded Latent Bridge for Heterogeneous LLM Communication
NeurIPS 2026 · Table 1
LLM agents
778.18Causal Prompting
LLaMA3
Causal Prompting: Debiasing Large Language Model Prompting Based on Front-Door Adjustment
AAAI 2025 · Main results table
Language, safety and interpretability
877.36SiReRAG (GPT-3.5-Turbo)SiReRAG: Indexing Similar and Related Information for Multihop Reasoning
ICLR 2025 · QA performance table
Retrieval and knowledge
976G-reasoner
GPT-4o-mini
G-reasoner: Foundation Models for Unified Reasoning over Graph-structured Knowledge
ICLR 2026 · Table 2
Retrieval and knowledge
1075.5HippoRAG 2
Llama-3.3-70B-Instruct
From RAG to Memory: Non-Parametric Continual Learning for Large Language Models
ICML 2025 · QA performance table
Retrieval and knowledge
49 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

HotpotQA · accuracy

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

#accuracy ↑MethodPaperTopic
190VUD AU filtering
Qwen2.5-14B-Instruct
Variational Uncertainty Decomposition for In-Context Learning
NeurIPS 2025 · Abstention on QA tasks table
Uncertainty and Gaussian processes
289.3MAS2
Qwen3-8B meta-agents
MAS$^2$: Self-Generative, Self-Configuring, Self-Rectifying Multi-Agent Systems
ICLR 2026 · Table 1
LLM agents
388.2KAR3
Llama-3.1-70B-Instruct
From Complex to Atomic: Enhancing Augmented Generation via Knowledge-Aware Dual Rewriting and Reasoning
ICML 2025 · Table 2
Retrieval and knowledge
488DCI-Agent-CC
Claude Sonnet 4.6
Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction
NeurIPS 2026 · Table 1
Retrieval and knowledge
585.8DTKG
Llama 3 8B
DTKG: Dual-Track Knowledge Graph-Verified Reasoning Framework for Multi-Hop QA
ICML 2026 · Table 1
Retrieval and knowledge
682.57QueryAligner
gpt-4o-mini
QueryAligner: Customizing User Query to Match LLMs Preferences for Better Intent Recognition
AAAI 2026 · Table 1
Alignment and preferences
781.1MoT (k=4, T=2)
GPT-4o-mini
Mixture-of-Trees: Learning to Select and Weigh Reasoning Paths for Efficient LLM Inference
AAAI 2026 · Table 2
Reasoning and chain of thought
879.7HiVA
Qwen2.5-72B-Instruct-Turbo
HiVA: Self-organized Hierarchical Variable Agent via Goal-driven Semantic-Topological Evolution
AAAI 2026 · Table 1
LLM agents
978.33ParamAgent
Llama-3.1-8B
ParamMem: Augmenting Language Agents with Parametric Reflective Memory
ICML 2026 · Table 1
Agents, code and math
1077.9EvoMAS
GPT-4o-mini
EvoMAS: Heuristics in the Loop—Evolving Smarter Agentic Workflows
ICML 2026 · Table 2
LLM agents
28 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.