Explore
Everything AION read, in seven sections. Pick one, a topic or a time window.
VideoSame Model, Different Speed: Why Your Inference Provider Matters — FriendliAI
Open-weight models are good enough now.
.png)
I Made Terrible Games With Google’s AI Playground
A long day’s haul in the video game slop mines.
DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training
We present DreamTrue, a multi-view, cross-embodiment robot world model for action-faithful and physically plausible video prediction.
Pollo AI turns creative ideas into campaigns with OpenAI
With GPT-5.6, GPT-6 Astra, and GPT‐Image‐2.5, Pollo AI helps creators turn bold ideas into detailed images and cinematic video ads.
Which Skill to Distill? SGUID: Selecting a Compact Skill Bank for Model-Skill Co-Evolution
Skills, reusable procedural guidance added at inference, can substantially improve LLM downstream performance (Li et al., 2026).
Show HN: Jevman – AI decision models play Pac-Man
Openai just launched their decisions endpoint, cloudflare launched clef the other week, and many more jev alternatives are out there.
Google Research RRSI Guide: Mastering Self-Improving AI Agents
In this tutorial, we implement RRSI (Regularized Recursive Self-Improvement), a method that lets an LLM agent rewrite its own harness, prompts, tools, memory, control flow, and sub-agents around a frozen model, without the harness overfitting to the tasks it evolves on.
unslothai/unsloth v0.1.905-beta: Sandboxing is here!
We're introducing Windows, Mac and Linux sandboxing in Unsloth!
anthropics/claude-code v2.1.294
Fixed prompt and agent hooks written as instructions (such as "Block commands that...") allowing what they should block
Synthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
We go deep on Periodic’s vision for “synthesis superintelligence”: reinforcement learning grounded in physical experiments, AI-powered materials characterization, simulations and density functional theory, high-throughput labs, and systems that learn from the entire process of doing science rather than only its published results.
Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses
Harnessed Agentic RL: Microsoft Research Asia introduces a training paradigm in which the same agent harness used in deployment participates directly in reinforcement learning, removing the need to reimplement the agent inside the training framework.
PaperRISED: Rubrics for Agentic Multi-Environment Selection and Self-Distillation
RISED: Rubrics for Agentic Multi-Environment Selection and Self-Distillation
One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO
Starting from Nemotron 3, our teams used supervised fine-tuning (SFT), reinforcement learning (RL), and feedback-driven inference to create systems that reached gold-medal level at both IMO 2026 and IOI 2026.
huggingface/trl v1.14.2
Patch release fixing two cases of silently wrong training and three crashes.
VideoTeaching Agents to Search with NVIDIA Data Designer — Dhruv Nathawani, NVIDIA
Dhruv Nathawani uses that funnel to explain how NVIDIA builds synthetic data that teaches a model to search, rather than answer from memory.
VideoBuilding Self-Learning Loops for Your Agent — Fuad Ali, Arize AI
A shopping assistant confidently returns products outside a user's budget because its price filter is broken.
MiMo-V2.6: Scaling Reinforcement Learning Towards Self-Improvement
Reinforcement learning (RL) is the central training paradigm for advancing large foundation models towards self-improvement.
A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization
We therefore introduce AgenticBBO-Bench, a cross-domain benchmark for agentic BBO spanning synthetic functions, hyperparameter optimization, database tuning, chip design, and molecular design under a unified finite-budget evaluation protocol.
ViSkill: Reinforcing VLM Agents with Evolving Visual-Native Skills
We propose ViSkill, a visual-native skill learning framework that encodes successful interactions as composite visual skill cards directly accessible to VLM agents.
Q-Learning with Scalar Adjoint Matching
Adjoint matching offers a principled way to update the flow model itself by propagating value information from the final action back to each flow step, but it requires a vector--Jacobian product through the policy at every step, a cost that grows with the number of flow steps and the policy size.
Learn2Play Bench: How Well Do LLM Agents Learn from Experience in Unfamiliar Environments?
Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments.
Composing What Each Teacher Learned: Multi-Teacher On-Policy Distillation through Teacher-Relative Shifts
We introduce Δ-MOPD, which transfers each teacher's teacher-minus-base logit shift re-anchored at the student's frozen initialization, and compare it with endpoint supervision in both settings while holding teacher selection fixed.
MIMESIS: Learning User Simulators as Training Environments for Interactive Agents
We introduce MIMESIS, a purpose-built user simulator trained on human conversations with explicit reasoning supervision and 13 realistic behavioral patterns derived from real user interactions.
MetaOPD: Meta-Learned Token Weighting for On-Policy Distillation
In this paper, we propose MetaOPD, a bilevel optimization framework that jointly learns the student model and a lightweight token-weighting network.
Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight
Reinforcement learning with verifiable rewards (RLVR) turns agent experience into learning signals primarily through scalar outcome rewards after interaction.
Opera: A Verbal Critic Framework for Long-horizon Coding Agents
We present Opera, a verbal critic framework that treats each correction as a persistent note, followed until the diagnosed problem is resolved.
Residual Advantage: Student-Relative Teacher Guidance for RL with Verifiable Rewards
We propose Residual Advantage (\RA{}), which treats the teacher--student probability residual as a bounded one-step reward, subtracts the corresponding state value under the student policy to form a standard advantage, and centers the result within each response before adding it to the verifier advantage.
Can AI Agents Learn Their Way to the Top? Evaluating Heuristic Learning in a Long-Running Game Agent Competition
Adversarial games have driven advances from heuristic search to reinforcement learning, yet learning and adapting strategies from limited samples remain challenging.
UniSkill: Learning Actor-Aligned Skill Proposals for an Evolving Policy
In this paper, we introduce UniSkill, which uses a shared policy to interact with the environment and propose skillbank edits (Add, Update, or No Edit) from the resulting trajectories.
Cost-Aware Mixture-of-Experts Coordination for Model Markets
This paper proposes an MoE-based model market framework that lifts Mixture-of-Experts from a model-level learning architecture to a market-level coordination mechanism.
System Switch: When Should a Fast Decision Model Stop and Think?
Dual-process agents pair a fast policy with a slow deliberative model.
CausalDreamer: Learning Predictive World Models with Latent Disentanglement
We propose CausalDreamer, which keeps the tokenizer frozen and re-encodes its latent into a factored representation of four groups along two axes: controllability, where only the two controllable groups receive the action, and reward relevance, learned by predicting the reward from the two reward-relevant groups.
On the Clock: Towards Punctual and Productive Time-Budgeted AI Agents
We study whether small LLM agents can operate effectively under explicit wall-clock time budgets by both respecting the allocated runtime and using available time productively.
Safe Meta-Policy Design with Risk Control
We study how to plan policy updates (i.e., meta-policy) before future candidates are trained, balancing the benefits of improvement against the risk of performance regression.
CAPABLE: Capability-Aware Policy Adaptation via Behavioral Latent Encoding
We introduce CAPABLE, a unified capability-aware adaptation framework for frozen VLAs that integrates self-supervised capability inference with residual reinforcement learning.
Radisson Hotel Group brings hotel discovery into ChatGPT
Radisson partnered with Accenture to build a ChatGPT plugin using OpenAI technology, helping travelers find, compare, and book hotels while planning their trips.
PIVOT: Perplexity-Informed KD-to-RL Transition Scheduling for Vertical-Domain Few-Shot Distillation
We propose PIVOT (Perplexity-Informed Transition Optimization), a dynamic transition framework that routes samples between OPD and GRPO according to teacher-evaluated sequence perplexity.
Fed-GRPO: Reward-Signal-Driven Federated Group Relative Policy Optimization
To this end, we propose Fed-GRPO, a federated GRPO training framework that addresses these privacy constraints by enabling collaborative reasoning training without sharing raw data, which leverages the reward statistics naturally produced during GRPO training as zero-cost signals to guide aggregation, local training, and communication.
Recursive Self-Improvement through Multi-Agent Self-Supervision
To address this, we propose Multi-Agent Self-Supervision (MASS), an RSI method that alternates between evolutionary workflow optimization and supervised fine-tuning on self-generated trajectories.
DIAL-OPD: Learning More from Fewer Tokens in On-Policy Distillation
We propose DIAL-OPD, a token-selection method that bridges log-probability and probability spaces by weighting reward magnitude with the logarithmic mean of teacher and student probabilities.
Agent Plasticity: Measuring Self-Improvement Through Experience
We introduce agent plasticity, the efficiency with which an agent converts experience into gains in future held-out performance.
AdaptEvo: Adaptive Agent Learning with Evolving Supervision
To address these challenges, we introduce AdaptEvo, a framework for learning under imperfect supervision that couples confidence-adaptive policy optimization with evolving decision knowledge and evaluation rubrics.
RL-ARC: Calibrating Large Reasoning Models via Reasoning-guided Uncertainty
To this end, we propose RL-ARC, a calibration-aware training framework that jointly leverages reasoning confidence and answer confidence.
unslothai/unsloth v0.1.904-beta: Train your own Decision model
Turn any text or vision LLM into a Jev-style decision model in Unsloth, with decision accuracy going from 30% to 80%.
Policy Alignment: New Signals for Membership Auditing in On-Policy Distillation
In this paper, we propose Policy Alignment Membership Auditing (PAMA), a new auditing framework tailored for OPD.
Experience-Guided Initiation Search for Learned Skills in Skill Composition
We propose EVIS, an Experience-Guided and Behavior-Validated Initiation Search framework for discovering reliable initiation configurations under limited target interaction.
Constitutional Gating and Deterministic Recovery for Multi-Agent LLM Negotiation: Ablations Against a Stateful Adversarial Gatekeeper
We study a three-part control stack - a 5-Pillar runtime constitution, a 4-tier swarm (Director, three-agent majority vote, Monitor, schema hard gate) and Cognitive Annealing (deterministic deadlock detection, atomic purge of the agent-side context, a canonical recovery message) - against a released adversarial Gatekeeper whose acceptance rules are fixed regular expressions and whose LLM only renders reply text.
Constrained Command-Conditioned Reinforcement Learning with Bandit Strategy Selection in Real-Time Strategy Games
Deep reinforcement learning agents reach strong performance in real-time strategy games but can be brittle against opponents outside their training distribution.