adaptive curriculum
Imagine a teacher who only gives you problems at the edge of your ability, neither too easy nor too hard. That is an adaptive curriculum in machine learning: training examples are automatically chosen based on the model's current performance to keep learning efficient. The focus shifts to tasks where the model is struggling but can improve, rather than wasting compute on solved or impossible ones.
The papers demonstrate that adaptive curricula consistently outperform fixed schedules. Methods include dynamically selecting puzzle difficulties, balancing rewards from test suites by difficulty, and using bandit algorithms to pick failure-prone scenarios. This approach lifts models on hard benchmarks, often closing gaps to larger systems or enabling improvement without external data. The key is prioritizing the frontier of capability.
- Papers
- 26
- Released code
- 1
- First seen
- Aug 2026
- Latest
- Nov 2026
23 papers in the last two months, against 3 in the two before.
Start here
Three papers to read first, in the order they came out- 1Adaptive test weighting gives software agents denser, more useful rewards
CodeLift dynamically calibrates test pass-rate rewards to create an adaptive curriculum for training SWE agents.
Oct 2026 - 2Targeting uncertain local decisions cuts GUI agent training cost by 56-61%
A two-stage adaptive training framework for GUI agents reduces environment interaction by 56-61% while improving task performance.
Oct 2026 - 3Adaptive curricula in AI agents secretly degrade the learner's ability to keep training
Unsupervised Environment Design fails when learners lose trainability due to parameter growth and degraded representations.
Nov 2026
How fast it is moving
Often studied with
reinforcement learningA machine learning method where an agent learns to make sequences of decisions by receiving rewards or penalties for its actions.6 togetherco-evolving feedbackFeedback that evolves alongside the policy or components it guides to improve performance.6 togetherself-evolving agentAn agent that improves by reusing its own experience, memory, or self-generated feedback, with success depending on verification and design.6 togetherWho is working on it
The papers
Most central to this idea first, not most recent.
- Big Techcs.LG
Adaptive curricula in AI agents secretly degrade the learner's ability to keep training
Unsupervised Environment Design fails when learners lose trainability due to parameter growth and degraded representations.
Mila – Quebec AI Institute, Université de Montreal · Nov 2026
- Chinese Techcs.AI
Targeting uncertain local decisions cuts GUI agent training cost by 56-61%
A two-stage adaptive training framework for GUI agents reduces environment interaction by 56-61% while improving task performance.
Alibaba Group, The Chinese University of Hong Kong · Oct 2026
- Chinese Techcs.LG
Adaptive test weighting gives software agents denser, more useful rewards
CodeLift dynamically calibrates test pass-rate rewards to create an adaptive curriculum for training SWE agents.
The Chinese University of Hong Kong, Shenzhen, Tencent · Oct 2026
- Big Techcs.AI
Adaptive training curricula produce stronger optimized agent harnesses
ActiveSaddler adaptively selects training scenarios for harness optimization, raising Pass@1 by 4.4 and 7.5 points on GAIA2 and Terminal-Bench 2.0.
POSTECH, KAIST · Oct 2026
- Big Techcs.RO
Adaptive sampling helps million-environment robot training solve harder tasks
Sampling task configurations where a policy has moderate success rates lets massively parallel robot RL solve harder tasks.
University of Washington, NVIDIA · Oct 2026
- Chinese Techcs.AI
Adaptive curriculum training lets vision-language models solve hard visual Sudoku puzzles
Adaptive curriculum learning via online rollout accuracy enables RLVR to solve long-horizon visual Sudoku puzzles where standard methods fail.
Peking University, Huawei Technologies Co., Ltd. · Oct 2026
- Big Techcs.AI
Adaptive instruction selection helps language models learn from peers
Adaptive instruction selection via a Stackelberg game improves multi-LLM collaborative alignment across benchmarks.
University of Washington, Amazon · Oct 2026
- Big Techcs.AI
Graded curricula extracted from an environment's latent rules substantially improve trained agents
Gradus composes guarded behavioral operators into calibrated curricula, lifting a 30B model from 59% to 89% on tau2-Bench and 47% to 83% on CAR-Bench.
AWS AI, University of Southern California · Oct 2026
- Big Techcs.AI
Adaptive judges help embodied agents improve as failure patterns change
VeriFine co-evolves policy, curriculum, and judge to scale verification and enable continuous self-improvement in embodied reasoning.
NVIDIA, UCLA · Oct 2026
- Top Universitycs.AI
Self-play on generated programs produces transferable language-model capabilities
Self-play pretraining with synthetic data from a universal Turing machine yields predictable zero-shot scaling on natural data.
Independent Researcher, Tel Aviv University · Sept 2026
- Industrycs.AI
Robots can act at 50 Hz while updating visual understanding once a second
Leap-VLA decouples slow semantic from fast reactive control, reusing cached VLM context to match same-frequency performance.
Leapmotor · Nov 2026
- Big Techcs.SE
Terminal agents can build adaptive environments for their own training
AutoEnvScaling uses terminal agents to automate RL environment generation, improving performance and enabling recursive self-improvement.
Northeastern University, Microsoft · Oct 2026
- Big Techcs.CV
Co-evolving visual tasks keeps model training near its learning frontier
Co-evolving visual environments with vision-language models via rollout pass rates improves out-of-domain reasoning by up to 5%.
Virginia Tech, Cisco · Oct 2026
- Big Techcs.AI
Prior barriers quantify uneven pretrained concept support for fine-tuning
Prior barriers in pretrained models follow a long-tail distribution, requiring adaptive instruction selection for effective fine-tuning.
Virginia Tech, Amazon · Oct 2026
- Big Techcs.AI
Self-evolving visual agents learn tool use from unlabeled images using margin reward.
A margin reward based on the solver's performance difference with and without tools enables visual agents to learn tool invocation from unlabeled images.
UCAS, Microsoft Research Asia · Oct 2026
- Top Universitycs.LG
Fisher-Rao prompt curricula make GRPO training more rollout-efficient
ARCUS reduces GRPO rollouts by 48-57% and improves accuracy by 2.8-2.9 points using Fisher-Rao arc length for prompt selection.
University of Wisconsin–Madison, University of Washington · Sept 2026
- Big Techcs.LG
MILO co-evolves agent harnesses and the strategies that discover them
MILO co-evolves agent harnesses and search strategies via meta-evolutionary island orchestration, outperforming state-of-the-art methods.
Georgia Institute of Technology, AWS AI Labs · Oct 2026
- Big Techcs.RO
Simulation-first framework lets coding agents learn real robot skills in 10 minutes
SimEX lets coding agents learn real robot skills by first practicing in simulation and then adapting with only 10 minutes of physical trials.
Amazon FAR, University of Texas at Austin · Oct 2026
- Chinese Techcs.AI
Adaptive feedback helps reinforcement-learning agents target their changing weaknesses
ARISE adapts evaluation criteria and training priorities to an agent's evolving capabilities, improving long-horizon task performance.
ShanghaiTech University, Ant Group · Sept 2026
- Independentcs.AI
AI models improve at predicting physics when objects are removed
VLMs trained on self-proposed, simulator-verified counterfactual physics tasks improve causal physical reasoning and transfer zero-shot to unseen dynamics.
Oct 2026
- Big Techcs.CV
Adaptive feedback targets each image model’s changing weaknesses
Adaptive, feedback-driven post-training improves diverse text-to-image models more than handcrafted reinforcement-learning data by targeting each model’s evolving weaknesses.
University of Science and Technology of China, Microsoft Research Asia · Oct 2026
- Chinese Techcs.AI
Evolving terminal environments keeps training tasks challenging as agents improve
Environment evolution off-policy generates increasingly difficult terminal environments, improving agent performance by 14-18 percentage points.
Hunyuan Team, Tencent · Sept 2026
- Chinese Techcs.AI
AI web agents learn long tasks by reusing skills salvaged from failures
SCOPE achieves 3.9% to 18.6% relative gains on web tasks by progressively discovering and composing local skills.
Renmin University of China, Huawei Technologies · Nov 2026
- Top Universitycs.AI
AutoSciBench automatically generates scientific agent benchmarks that adapt as agents improve.
AutoSciBench generates and refines scientific-agent benchmarks, producing harder and higher quality tasks than human-curated ones.
KAIST, Genentech · Oct 2026
- Top Universitycs.ROcode
One video world model transfers physical dynamics across robot bodies
A single video model trained on diverse human and robot data can simulate physics for unseen embodiments zero-shot.
Princeton University · Aug 2026
- Independentcs.AI
Five-level ladder maps reasoning models beyond direct human oversight
Autonomous rewards and self-generated experience enable large reasoning models to improve beyond human supervision.
Sept 2026
Concepts are extracted from each paper and reused across the corpus, so this page grows on its own as the desk reads.