ARC

The AI2 Reasoning Challenge — grade-school science multiple-choice questions split into an Easy and a harder Challenge set.

Papers using ARC (20)

Improved Large Language Diffusion Models2026Prompt Robustness Is Task-Dependent: Comparing Objective and Belief-Style Questions in LLM Evaluation2026Pruning via Causal Attribution Preserves Reasoning Performance in Large Language Models2026Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injection2025 · 1 citesTurning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injection2025 · 1 citesQEIL v2: Heterogeneous Computing for Edge Intelligence via Roofline-Derived Pareto-Optimal Energy Modeling and Multi-Objective Orchestration2026Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs2026Data-Free Pruning of Self-Attention Layers in LLMs2025Uncertainty-Aware Answer Selection for Improved Reasoning in Multi-LLM Systems2025Dr.LLM: Dynamic Layer Routing in LLMs2025Uncovering Cross-Linguistic Disparities in LLMs using Sparse Autoencoders2025Text-to-LoRA: Instant Transformer Adaption2025From Threat to Tool: Leveraging Refusal-Aware Injection Attacks for Safety Alignment2025Temporal Self-Rewarding Language Models: Decoupling Chosen-Rejected via Past-Future2025LLMs Can Get "Brain Rot"!2025More is Less: The Pitfalls of Multi-Model Synthetic Preference Data in DPO Safety Alignment2025Order Independence With Finetuning2025ORI: O Routing Intelligence2025Teuken-7B-Base & Teuken-7B-Instruct: Towards European LLMs2024 · 2 citesLanguage Models are Hidden Reasoners: Unlocking Latent Reasoning Capabilities via Self-Rewarding2024 · 1 cites