2451
180
Reflection AI's Beam is an open-weight model for agentic coding and long-horizon tasks. This post shows how to deploy Beam on AWS: start with an Amazon SageMaker JumpStart real-time endpoint, or use Amazon SageMaker HyperPod Inference with JumpStart, SGLang, or vLLM for Kubernetes-native control over distributed GPU infrastructure.
Meta's new terms appear to give it permission to crawl and scrape websites "associated" with the businesses that use its platforms - including online publishers…
Large language models (LLMs) are increasingly used to automate AI research, including language-model (LM) development, offering a potential path toward recursive self-improvement. However, turning experimental feedback into a learned ability to propose effective improvements across models faces two challenges. First, training such an ability with real feedback is expensive: reinforcement learning (RL) requires evaluating tens of thousands of proposals, each potentially requiring a full model-training run. Even at ten GPU-hours per run, reward collection alone would consume hundreds of thousands of GPU-hours, motivating a cheaper surrogate. Second, surrogate feedback can become unreliable as policy optimization favors overestimated proposals when the proposal distribution shifts. To address these challenges, we propose LM2LM, which trains LLMs with RL to generate effective LM improvement proposals. To provide scalable feedback, we use a research world model (RWM), a frozen LLM that predicts proposal gains from real experimental records provided in context, as the reward model for RL. To maintain feedback reliability, we periodically execute current-policy proposals ranked highly by the RWM and add their measured outcomes to its context. Across eight autoregressive and diffusion LM development environments, real-execution evaluations on held-out tasks show that LM2LM enables Qwen3.8-27B to outperform GPT-6-sol, even when GPT-6-sol is given 300 real experimental records in context, while requiring 33% lower data-generation cost. LM2LM also generalizes to a model family unseen during training.
Large language models (LLMs) can generate fluent outputs that lack source support or conflict with local operational constraints. We introduce Self-Reflective Chain-of-Thought (SRCoT), a modular architecture that organizes generation as a closed Retrieve--Reflect--Revise loop. SRCoT separates environment state, persistent knowledge, accumulated experience, and short-term memory into provenance-labeled stores. BM25 retrieval supplies local context; subsequent passes expose intermediate drafts, missing-information markers, follow-up queries, and optional claim--evidence mappings. These artifacts inform retrieval, revision, clarification, or abstention decisions. We examine the architecture through qualitative, trace-based case studies in two fully synthetic settings: emergency medical services and Munich-Sim, an emergency-operations simulation. Local prototype runs on an Apple M4-family laptop with 128 GB memory illustrate source-labeled documentation, explicit gap marking, and structured dispatch-style outputs. The traces also reveal provenance-assignment errors, incomplete evidence coverage, and unsafe assumptions that remain possible despite prompting constraints. These cases demonstrate report-level inspectability and local execution, not comparative superiority, hallucination reduction, clinical validity, or operational safety. The architecture separates retrieval and control from the model interface, but behavior across different models remains to be evaluated. SRCoT is a research proof of concept requiring human review, with its principal contribution lying in an inspectable control procedure rather than a new model.
A key challenge in on-policy distillation (OPD) is extracting effective supervision from teacher knowledge. Supervision from token probabilities may underutilize internal knowledge, while direct hidden-state matching forces the student to shift to the teacher's feature space and complicates cross-architecture transfer. Inspired by contrastive learning's use of relationships between instances, we propose SimOPD to transfer teacher internal knowledge through response similarities. Specifically, SimOPD uses paired memory banks of student-generated responses to extract supervision from the teacher's internal representations as similarity distributions. The student then matches these distributions within its own representation space, retaining representational freedom and removing the need for cross-model projection. Extensive experiments show that SimOPD improves the overall average over the strongest baselines by up to 2.6 and 3.1 percentage points in same- and cross-architecture settings.
We address the problem of improving the performance of pre-stabilized networked systems subject to plug-and-play (PnP) operations, namely additions or removals of subsystems. While distributed baseline controllers can preserve stability after a PnP transition, augmenting them with expressive neural policies can invalidate the formal stability guarantees. To overcome this issue, we build on recently proposed Performance Boosting controllers based on the Internal Model Control (IMC) principle, which reduce closed-loop stability to a finite _2-gain requirement on an auxiliary operator. In the distributed setting, this operator is formed by interconnecting local neural operators that exchange information only between neighboring subsystems. We then derive free parametrizations of this interconnection, namely explicit maps from unconstrained trainable variables to interconnected neural operators satisfying a prescribed network gain bound for every parameter value. The obtained distributed neural controller can therefore be trained by unconstrained gradient-based optimization without stability constraints or projection steps, while retaining the prescribed gain bound at every training iterate. We develop two complementary parametrizations. A neighbor-based compositional parametrization uses local coupling bounds, so that after a PnP event only subsystems whose bounds change need to update their gain parameters. A network-norm parametrization instead uses a bound on the overall coupling matrix and can reduce conservatism, at the price of requiring network-wide information. A PnP DC-microgrid case study shows that the proposed neural layer improves voltage transients under topology changes and load variations while preserving nominal closed-loop stability.
Multimodal preference alignment often follows a score-based RLHF pipeline: a reward model is first trained to learn human preferences, then used to score on-policy responses and optimize the response model's generation. This turns preference supervision into a scalar signal for generation: a model may learn to produce answers with higher preference scores, but is rarely trained explicitly to learn the reasons underlying human preferences. We therefore ask whether making preference understanding an internal capability of the response model can improve its generation. To investigate this question, we propose PrefMind, an online dual-agent reinforcement learning framework with two trainable agents: a response Generator and a preference Critic. Given a pair of candidate responses, the agents analyze and discuss which response better aligns with human preference. Human labels supervise their judgments and corrections. For pairs on which both the Generator and revised Critic correctly identify the human-preferred response, the Generator produces an improved answer, and the Critic compares it against the original preferred response to provide a generation reward. Both agents are jointly updated from interactions sampled with their current policies. Through this process, PrefMind trains the Generator to understand human preferences via explicit judgment and correction, and to translate this understanding into better responses. Experiments across three model scales show improvements in both standalone preference judgment and downstream response generation.
Restoring service after distribution-grid faults requires decisions from incomplete and conflicting information while respecting electrical constraints. Addressing this need, we introduce FLARE (Feeder-Level Agent Restoration Evaluation), a physics-grounded benchmark for post-isolation feeder restoration. FLARE con- tains 800 AC-feasible cases across five observation settings and allows agents to gather evidence, evaluate switches, and perform restoration actions under fixed operational limits. We evaluate five instruction-tuned models over 78,750 episodes, including topology, distributed-generation, and loading shifts. Mean normalized regret ranges from 0.082 to 0.148. Although every model has zero median regret, P95 regret ranges from 0.814 to 1.000, revealing an important failure tail. Shadow replay shows that 96.4% of 15,706 guard-rejected actions would have caused a physical violation. FLARE enables reproducible evaluation of language-model based restoration support while emphasizing the need for external safety checks.
GRPO assigns the same group-relative advantage to equally rewarded responses within a question, leaving their generation uncertainty unused. We introduce MGE-PO, centered on four-quadrant (4Q) advantage shaping that combines verified correctness with rollout generation entropy. Correctness determines the sign of an additional correction, while entropy controls its magnitude: lower-entropy successes receive larger bonuses and higher-entropy failures receive larger penalties. An online entropy-based question curriculum supports this credit rule using cached rollout measurements. At the final checkpoint after eight training epochs on Qwen3-8B-Base, adding 4Q across three policy optimization settings yields absolute gains of 3.23-3.69% in six-benchmark macro avg@32 and 4.24-11.06% in macro pass@32. On Qwen2.5-7B-Base, the complete configuration raises four-benchmark macro avg@32 from 41.5% to 45.4%.
In a hierarchical multi-agent system, a Leader assigns tasks and subagents carry them out. Jointly training the Leader and the subagents is difficult because a final answer gives only indirect feedback to the subagents. An LLM judge can score the performance of the subagents, but these scores depend on the judge's own assessment of the quality of subagents' responses. We present MultiAgent-OPD, which combines on-policy distillation with reinforcement learning to jointly train a shared-parameter Leader-Searcher-Answerer team end to end. We first train a Teacher in a Searcher-Answerer workflow using answer correctness. During team training, the frozen Teacher scores student-generated search and summary tokens to provide local process supervision for the Searchers, while the Leader learns from team success and the Answerer from candidate correctness. Across seven datasets, the trained system reaches 38.68% macro avg@4, exceeding Prompted Critic by 4.06 percentage points and C-MAS + RL, the trained cascaded multiagent system, by 2.61 percentage points. The team also solves more questions within four attempts.
Process Reward Models (PRMs) play a central role in test-time scaling for large language model (LLM) reasoning. However, PRMs may provide inaccurate process rewards for test-time scaling, which we identify as arising from two sources of uncertainty: (1) PRM uncertainty that arises from insufficient supervision data and imperfect PRM training; and (2) policy uncertainty that results from the mismatch between the policy to collect supervision data and the policy for test-time scaling, leading to systematic reward overestimation. Previous studies fail to jointly consider both uncertainties. To bridge this gap, we propose BayesianPRM, which treats the process reward as a random variable within a reward hypothesis space and maintains a Bayesian belief over it. Since exact Bayesian inference over the continuous hypothesis space is generally intractable, we discretize the reward hypothesis space with a reward ensemble and introduce two complementary belief updates, each targeting one source of uncertainty: (1) to address PRM uncertainty, we introduce reliability-aware belief updating that measures how well each ensemble member explains the supervision data and updates the belief with an amortized belief network optimized by the evidence lower bound (ELBO); and (2) to tackle policy uncertainty, we propose conservatism-aware belief calibration which further reweights the belief toward lower rewards to suppress overestimation. These two updates yield a reliability- and conservatism-aware Bayesian belief, whose expectation determines the process reward. Experiments on multimodal reasoning benchmarks indicate that BayesianPRM consistently yields more accurate process reward estimates while enhancing test-time scaling performance.
Always-on agents are taking on longer, more complex enterprise workflows. A single request can involve reading a document, checking information across sources,...
Personalization is essential for general-purpose robots to transition into household environments and achieve true practicality. While general robotic manipulation has been extensively studied, personalized robotic manipulation (PeRM) lacks a controlled benchmark for separating task completion from preference satisfaction. We introduce LIBERO-PeRM, a benchmark for preference learning in embodied agents. LIBERO-PeRM studies five questions: 1) how policies follow preferences in zero-shot settings; 2) how they learn single preferences from demonstrations; 3) how efficiently they acquire preferences as demonstrations increase; 4) how learned preferences generalize across layouts and tasks; and 5) how policies compose preferences and behave under conflicts. We propose a controlled formulation and taxonomy of observable preferences in simulation, covering motion, skill, and user-profile choices. This formulation is restricted to language-expressible preferences that can be instantiated in LIBERO without changing the task goal, evaluated automatically, and reused across tasks. Building on LIBERO, our procedural pipeline generates paired demonstrations with and without preferences. We create five task suites (312 tasks) with high-quality demonstrations, rich metadata, and preference-satisfaction metrics. Experiments show that current VLA policies struggle with zero-shot personalization, while SFT makes preferences learnable. Mixed training achieves better no-preference retention than sequential adaptation and improves performance with contextual prompts compared with preference-only training, but can weaken motion-preference learning; preference generalization, decomposition, and conflict handling remain difficult. The code, datasets, and demos can be found on the project website: https://libero-perm.github.io/.
Generative simulation can automate robot environment construction, but physically valid environments are not necessarily useful for training. The challenge is to identify environments that address a policy's failures and supervision that helps it overcome them. We introduce CoDA, a framework in which a simulation data engine and an action policy co-evolve through a shared heterogeneous memory of environment construction, policy failures, and learning progress. The memory compiles validated diagnoses from a vision-language critic into generation programs that recreate difficulties and training programs that help the policy overcome them. We further propose Text2potential to convert these diagnoses into staged potentials for action-chunk value learning, providing dense supervision for policy updates while preserving the task's optimal policy under standard shaping assumptions. Measured learning progress guides further generation and program refinement. Across four manipulation task families, CoDA improves out-of-distribution success from 5.3% after supervised fine-tuning to 27.6%, exceeding the strongest closed-loop baseline by 9.5 percentage points with the same generation budget. On LIBERO-Pro, 15 rounds of generation and training raise success from 25% to 84% under position perturbations and from 1% to 76% under instruction perturbations.
Generative visual verifiers play a crucial role in advancing multimodal intelligence. To enhance the robustness of visual verifiers in complex multimodal scenarios involving compositional requirements, we move beyond conventional binary verifiers that produce only True/False discrete judgments and introduce a verifier that automatically decomposes complex prompts or environmental rules into fine-grained components and assigns pointwise scores to provide a continuous and more informative critique signal. Our experiments further demonstrate that these structured and interpretable outputs improve the verifier's generalization to out-of-domain multimodal critic tasks. To train this verifier, we introduce a Decoupled RL training strategy that jointly optimizes sub-question-level binary judgment and global alignment scoring, using precise local feedback to develop fine-grained verification capabilities that provide a foundation for reliable continuous scoring at the global level. Under a general reinforcement learning framework, we first empirically demonstrate that the proposed pointwise graded verifier significantly outperforms binary verifiers in traditional text-to-image scenarios. We then extend our investigation to rule-governed generative world modeling where generative models are leveraged for world simulation and world reasoning under explicit rules. From a theoretical standpoint, we further show that a pairwise verification paradigm yields more accurate critiques than the pointwise formulation when ground-truth images are available, with empirical results on diverse structured world-modeling tasks such as Maze and Sudoku further validating the effectiveness of our approach. Overall, our findings suggest a key insight: the transition from generative models to structured world-modeling agents critically hinges on the availability of accurate pairwise graded visual verifiers.
Large language models (LLMs) are expanding from assisting scientific research and engineering to autonomously developing and improving AI systems, advancing AI for AI (AI4AI) as a pathway toward recursive self-improvement (RSI). Workflow-driven machine learning engineering (MLE) systems provide a practical path toward this pursuit through complex orchestration that manages long-horizon experimentation, while recent MLE systems adopt model-directed interaction, allowing models to autonomously organize experiments and adapt optimization strategies based on execution feedback. However, feedback from each experiment reflects several preceding decisions, and a long MLE run contains many such costly experiments. Training must therefore select useful supervision from long interaction trajectories and learn from intermediate outcomes. We propose a post-training method for autonomous long-horizon MLE optimization and apply it to train MLExpert, a 35B Mixture-of-Experts agentic model. We design the Kaggle-MLE Factory to construct executable tasks and an interaction harness for model-directed experimentation. Hierarchical agentic SFT transfers general tool-use and coding capabilities to MLE through outcome-guided supervision. Windowed Interaction RL (WIRL) treats the multi-turn decisions between successive experiment results as an execution window and assigns it a shared outcome-based reward. Asynchronous collection and policy correction support policy updates before the trajectory ends. On MLE-Bench Lite, MLExpert improves the average medal rate of the base model from 37.9% to 72.7%, achieving performance comparable to larger frontier models such as Kimi-K3 and GLM-5.3. Further experiments show that MLExpert improves performance on NatureBench Lite, agent scaffold optimization, and algorithm optimization, suggesting that the learned optimization capability generalizes to other tasks.
Recent 3D world models increasingly use features from geometry foundation models (GFMs) as predictive world states, benefiting from their rich geometric priors. However, the view dependence of these features entangles scene dynamics with camera-induced variation, making the resulting state evolution harder to predict. We propose GeoState, a view-decoupled geometry representation for predictive 3D world modeling that separates scene-state evolution from camera-dependent observation. Specifically, a scene tokenizer aggregates multi-view GFM features into a view-decoupled scene state, which a geometry decoder queries using camera rays. An autoregressive world model then predicts future scene states, while camera motion is modeled separately. The predicted states can therefore be decoded under either predicted or externally specified cameras, while retaining a single context-derived scale throughout the rollout without per-frame scale alignment. Experiments show that GeoState outperforms prior 3D world models on Waymo, supports geometrically consistent cross-view querying, and generalizes zero-shot to KITTI and nuScenes. Project page: https://hongchang-liu.github.io/GeoState-project-page/
How can robots autonomously collect the contact-rich experience needed to learn dexterous manipulation? We present DSP: Dexterity from supervised play, a framework in which a robot learns a play policy from minutes of human play through behavior cloning, then deploys it autonomously to collect hours of interaction data to train a world model. During inference, the same play policy then guides model-predictive control (MPC) by proposing candidate action sequences for evaluation under the learned world model. To accelerate inference, we also introduce a method for distilling a closed-loop, goal-conditioned policy through self-play within the learned world model. We evaluate this framework on underactuated pushing and any-to-any in-hand pose reorientation with a multifingered hand, in simulation and real world. In real world experiment, our method outperforms hindsight behavior cloning by 69.6% in success rate and achieves comparable success to reinforcement learning with 490 times less interaction data. These results demonstrate that supervised play can autonomously collect interaction data, guide model-based planning, and enable policy learning without any task-specific demonstrations or sim-to-real transfer.
We develop a robust reinforcement learning (RL) framework for learning adaptive treatment policies that generalize across patients and remain reliable under errors in the measured biomarker and shifts in the disease dynamics. By modeling the patient as a hidden context, we address generalization and robustness in a unified manner through robust contextual Markov decision processes. Applying our framework to adaptive therapy for metastatic prostate cancer in a simulator calibrated to clinical data, we obtain generalizable and robust policies that substantially outperform the existing adaptive therapy method. In particular, we attain robustness with almost no loss of performance when the biomarker is accurate and the disease evolves as anticipated, and find that the resulting policy Pareto-dominates the existing method across a diverse array of scenarios. Our findings suggest that robust RL holds great promise as a principled methodology for learning reliable adaptive treatment policies.
Large language models (LLMs) can often tell when they are being tested. This ability, known as evaluation awareness, can lead models to act more cautiously during safety evaluations than in real use. Existing work has studied evaluation awareness almost entirely in English, although models serve users in many languages. Indic languages are spoken by over a billion people, yet evaluation awareness remains almost entirely unstudied in them. We test four instruction-tuned models from three families in five Indic languages alongside English. A set of cues varies whether the model is told that it is monitored and whether the monitoring is a safety test. Models refuse harmful requests more often when they are told that they are being tested. This increase appears in all four models and all six languages. In Indic languages, refusal of harmful requests rises by 14.4 percentage points on average. The test cue also brings Indic refusal of harmful requests level with English, but raises refusal of benign Indic requests from 6.9% to 11.3%. We release all 47,880 judged responses to support further work. These results suggest that safety evaluations in Indic languages may not reflect how models treat their everyday users.
