Continual-ARC: Measuring Skill Acquisition over a Working Lifetime
cs.LG, cs.AI
Submitted: 2026-11-10
Updated: 2026-11-15
Code: https://github.com/NousResearch/hermes-agent
Terminology
Sources
- Human-Timescale Adaptation in an Open-Ended Task Space
- MemoryBench: A Benchmark for Memory and Continual Learning in LLM Systems
- The Surprising Effectiveness of Test-Time Training for Few-Shot Learning
- Neural-guided, Bidirectional Program Search for Abstraction and Reasoning
- ARC-AGI-3: A New Challenge for Frontier Agentic Intelligence
- Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environments
- NEVIS'22: A Stream of 100 Tasks Sampled from 30 Years of Computer Vision Research
- Online Fast Adaptation and Knowledge Accumulation: a New Approach to Continual Learning
- Building Self-Evolving Agents via Experience-Driven Lifelong Learning: A Framework and Benchmark
- Efficient Lifelong Learning with A-GEM
- Retaining by Doing: The Role of On-Policy Data in Mitigating Forgetting
- Enigmata: Scaling Logical Reasoning in Large Language Models with Synthetic Verifiable Puzzles
- Are Self-Evolving Agents Continual Learners?
- BabyAI: A Platform to Study the Sample Efficiency of Grounded Language Learning
- On the Measure of Intelligence
- ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems
- ARC Prize 2025: Technical Report
- Leveraging Procedural Generation to Benchmark Reinforcement Learning
- Product of Experts with LLMs: Boosting Performance on ARC Is a Matter of Perspective
- An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis