AI Paper Atlas

Research map / Benchmarks / PopQA

PopQA leaderboard

38 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on PopQA. Best EM: 56 (Interact-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.

PopQA · EM

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

#EM ↑MethodPaperTopic
156Interact-RAG
Qwen3-8B
Interact-RAG: Reason and Interact with the Corpus, Beyond Black-Box Retrieval
ICLR 2026 · Table 1
Retrieval and knowledge
250.1ReSeek
Qwen2.5-7B-Instruct
ReSeek: A Self-Correcting Framework for Search Agents with Instructive Rewards
ICML 2026 · Table 3
LLM agents
350.1Search-R2
Qwen2.5-32B
Search-R2: Enhancing Search-Integrated Reasoning via Actor-Refiner Collaboration
ICML 2026 · Table 3
Reasoning and chain of thought
448.8SkillOrchestra+
Qwen2.5-3B
SkillOrchestra: Learning to Route Agents via Skill Transfer
NeurIPS 2026 · Table 1
LLM agents
547.9RAG-R1-mq
Qwen2.5-7B-Instruct
RAG-R1:Incentivizing the Search and Reasoning Capabilities of LLMs Through Multi-Query Parallelism
AAAI 2026 · Table 3
Retrieval and knowledge
646.9BAR-RAG (3 Iter)
Qwen-2.5-7B-Instruct
Rethinking the Reranker: Boundary-Aware Evidence Selection for Robust Retrieval-Augmented Generation
ICML 2026 · Table 1
Retrieval and knowledge
746.5InfoSeeker-3B
Qwen2.5-3B-Instruct
Enhancing Agentic Search via Data Synthesis on Hierarchical Constraint Satisfaction
ICLR 2026 · Table 3
Agents, code and math
845.2InForage
Qwen2.5-3B-Instruct
Scent of Knowledge: Optimizing Search-Enhanced Reasoning with Information Foraging
NeurIPS 2025 · Main experimental results tabl
Retrieval and knowledge
944.32SmartRAG
LlaMa-2 7B
SmartRAG: Jointly Learn RAG-Related Tasks From the Environment Feedback
ICLR 2025 · Table 1
Visual and representation RL
1044.3Stratified GRPO
Qwen2.5-3B-Instruct
Stratified GRPO: Handling Structural Heterogeneity in Reinforcement Learning of LLM Search Agents
ICML 2026 · Table 1
LLM agents
7 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

PopQA · accuracy

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

#accuracy ↑MethodPaperTopic
168.3RaCoT
Qwen2.5-7B
RaCoT: Plug-and-Play Contrastive Example Generation Mechanism for Enhanced LLM Reasoning Reliability
AAAI 2026 · Table 2
Retrieval and knowledge
266.2InstructRAG-FT
Llama-3-Instruct 8B
InstructRAG: Instructing Retrieval-Augmented Generation via Self-Synthesized Rationales
ICLR 2025 · Main results table
Retrieval and knowledge
362.3MemGen-GRPO
Qwen3-8B
MemGen: Weaving Generative Latent Memory for Self-Evolving Agents
ICLR 2026 · Table 1
LLM agents
460.6WebSeer (Local RAG)
Qwen2.5-14B
WebSeer: Training Deeper Search Agents through Reinforcement Learning with Self-Reflection
ICLR 2026 · Table 1
Human-AI and goal-conditioned
557.54Verifier-8x7B + Drafter-7B
Mistral-7B drafter; Mixtral-8x7B verifier
Speculative RAG: Enhancing Retrieval Augmented Generation through Drafting
ICLR 2025 · Main results table
Retrieval and knowledge
657.3Self-Reasoning 13B
LLaMA2-13B
Improving Retrieval Augmented Language Model with Self-Reasoning
AAAI 2025 · Table 1
Retrieval and knowledge
755.8Self-RAG 13B
Llama2-13B
Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection
ICLR 2024 · Table 1
Retrieval and knowledge
845AutoRefine-Base
Qwen2.5-3B-Base
Search and Refine During Think: Facilitating Knowledge Refinement for Improved Retrieval-Augmented Reasoning
NeurIPS 2025 · Table 1
Retrieval and knowledge
944.8C-3PO
Qwen2-72B
C-3PO: Compact Plug-and-Play Proxy Optimization to Achieve Human-like Retrieval-Augmented Generation
ICML 2025 · Main results table
LLM agents
1042.6SMART
Llama-2-7B
Synergistic Multi-Agent Framework with Trajectory Learning for Knowledge-Intensive Tasks
AAAI 2025 · Main results table
LLM agents
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

PopQA · F1

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

#F1 ↑MethodPaperTopic
171.16Sparse RAG
Gemini XXS
Accelerating Inference of Retrieval-Augmented Generation via Sparse Context Selection
ICLR 2025 · Quality & Efficiency Tradeoff
Retrieval and knowledge
257.6PANINI
GPT-4o-mini answer model
Panini: Continual Learning in Token Space via Structured Memory
ICML 2026 · Table 2
Retrieval and knowledge
356.74Psi-RAG (summative abstract)
Llama-3.3-70B
Hierarchical Abstract Tree for Cross-Document Retrieval Augmented Generation
ICML 2026 · Table 2
Human and embodied modeling
456.2HippoRAG 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
553.8IGPO
Qwen2.5-7B-Instruct
Information Gain-based Policy Optimization: A Simple and Effective Approach for Multi-Turn LLM Agents
ICLR 2026 · Table 1
RL from feedback
653.5DECOR
Qwen3-8B-Instruct
DECOR: Learning to Decompose and Collaborate in Deep Search via Multi-Agent Reinforcement Learning
ICML 2026 · Table 1
Cooperative multi-agent RL
751.22Graph-R1
Qwen2.5-7B-Instruct
Graph-R1: Towards Agentic GraphRAG Framework via End-to-end Reinforcement Learning
ICML 2026 · Table 1
Retrieval and knowledge
850.2JADE
Qwen2.5-7B-Instruct
JADE: Bridging the Strategic-Operational Gap in Dynamic Agentic RAG
ICML 2026 · Table 2
LLM agents
947.6SmartRAG
LlaMa-2 7B
SmartRAG: Jointly Learn RAG-Related Tasks From the Environment Feedback
ICLR 2025 · Table 1
Visual and representation RL
1046.2ReasonRAG (5k)
Qwen2.5-7B-Instruct
Process vs. Outcome Reward: Which is Better for Agentic RAG Reinforcement Learning
NeurIPS 2025 · Main Results table
Retrieval and knowledge
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.