Interactive benchmarks evaluate AI reasoning through dialogue, revealing hidden capabilities static tests miss.
Standard benchmarks have become increasingly unreliable due to saturation, subjectivity, and poor generalization. We argue that evaluating model's ability to acquire information actively is important to assess model's intelligence. We propose Interactive Benchmarks, a unified evaluation paradigm that assesses model's reasoning ability in an interactive process under budget constraints. We instantiate this framework across two settings: Interactive Proofs, where models interact with a judge to deduce objective truths or answers in logic and mathematics; and Interactive Games, where models reason strategically to maximize long-horizon utilities. Our results show that interactive benchmarks provide a robust and faithful assessment of model intelligence, revealing that there is still substantial room to improve in interactive scenarios. Project page: https://github.com/interactivebench/interactivebench
Interactive benchmarks measure AI reasoning by requiring models to actively query judges for information under budget constraints.
Models achieving zero percent accuracy on static tests unlock substantial reasoning ability when allowed interactive dialogue.
Interactive benchmarks reveal significant gaps in current AI evaluation methods that overlook information acquisition abilities.
Dataset of only 46 situation puzzles may be too small for comprehensive evaluation of model capabilities.
Interactive game benchmarks require environment setup and other agents, potentially limiting generalization to real-world deployment.
Meituan LongCat Team, Anchun Gui +160
We introduce LongCat-Flash-Thinking-2601, a 560-billion-parameter open-source Mixture-of-Experts (MoE) reasoning model with superior agentic reasoning capability. LongCat-Flash-Thinking-2601 achieves ...
Tianxin Wei, Ting-Wei Li +27
Reasoning is a fundamental cognitive process underlying inference, problem-solving, and decision-making. While large language models (LLMs) demonstrate strong reasoning capabilities in closed-world se...
Ailin Huang, Ang Li +213
We introduce Step 3.5 Flash, a sparse Mixture-of-Experts (MoE) model that bridges frontier-level agentic intelligence and computational efficiency. We focus on what matters most when building agents: ...
Veuns-Team, Changlong Gao +25
GUI agents have emerged as a powerful paradigm for automating interactions in digital environments, yet achieving both broad generality and consistently strong task performance remains challenging.In ...
Ji Zeng, Dayuan Fu +15
Recently, the frontier of Large Language Model (LLM) capabilities has shifted from single-turn code generation to agentic software engineering-a paradigm where models autonomously navigate, edit, and ...
