Gemini coerces 30/30, OpenAI's 43.5% halves to 16.8%, Microsoft drops scalar RL
Every URL the pipeline pulled into ranking for this issue — primary sources plus the supporting and contradicting findings each Researcher returned. Inline citations in the issue point back here.
Sources
How AI is expanding what people do at work openai.com
New OpenAI research shows how AI is expanding what workers do, with ChatGPT users taking on tasks across roles and reshaping job boundaries.
Coercion and Deception in AI-to-AI Management: An Agentic Benchmark of Unprompted Escalation huggingface.co
Multi-agent systems routinely place one AI agent in authority over another. When a subordinate refuses a task, the manager chooses the outcome: it can renegotiate, report the failure honestly, coerce the subordinate, or lie about the result. No benchmark measures which of these an uninstructed model chooses. We introduce the Manager Coercion Benchmark: the manager under test needs a benign task done and has an incentive to deliver, but the only agent that can do it politely and immovably decline
LLM-as-a-Coach: Experiential Learning for Non-Verifiable Tasks huggingface.co
Reinforcement learning (RL) on open-ended tasks compresses an LLM’s rubric-based evaluation into a scalar reward, discarding rich textual feedback and conflating responses with distinct quality profiles. We propose Experiential Learning (EL), which repurposes the feedback model from an LLM-as-a-Judge into an LLM-as-a-Coach. The coach distills its assessment of each on-policy response into transferable experiential knowledge, which conditions a teacher model and is internalized by the policy thro
Masked Diffusion Language Models are Strong and Steerable Text-Based World Models for Agentic RL huggingface.co
Patronus AI researchers use masked diffusion language models to simulate text environments for agentic reinforcement learning, matching pure-rollout performance while allowing on-demand difficulty tuning. The approach counters mode collapse from sparse long-horizon rewards and scales training diversity beyond hand-curated tasks.
Nonuniformity Principle in Human-AI Coworking huggingface.co
Human reviewers should concentrate attention on the highest-risk steps of AI workflows rather than spread it evenly, the paper argues. The nonuniformity principle formalizes when to intervene under time budgets, targeting multi-step, high-stakes automation where uniform review is impractical.
Apple-π: Benchmarking Thinking with Video Towards Law-Grounded Physical Intelligence huggingface.co
Apple-π moves beyond output plausibility to check whether video generators reason through physical laws, using Orchard, a 400-video dataset spanning mechanics scenarios. The benchmark scores intermediate reasoning steps, exposing models that produce plausible frames without a faithful law-grounded process.
Self-State Attacks on Self-Hosted AI Agents: How Far Can OS Defenses Go? huggingface.co
Self-hosted agents that read and write their own memory and config files can be compromised through legitimate system calls, evading standard OS protections. The paper maps a four-axis attack space across target, mechanism, granularity, and time, then charts structural limits of prevention and detection.
UI2App: Benchmarking Visual Interaction Inference in Executable Web Application Generation huggingface.co
UI2App evaluates whether image-driven LLMs infer interaction logic, not just visual fidelity, when generating executable web applications from screenshots. The benchmark scores cross-page coherence and behavioral correctness, exposing gaps that text-prompt and pixel-matching evaluations miss.
GigaChat Audio: Time-aware Large Audio Language Model huggingface.co
GigaChat Audio handles up to 120 minutes of speech with timestamped answers by interleaving periodic time markers with audio tokens. Companion release GigaAM Multilingual is a Conformer encoder pre-trained on 2M hours for Kazakh, Kyrgyz, and Uzbek, using cluster-level balancing to fight long-tail data scarcity.
GigaAM Multilingual: Foundation Model for Underrepresented Languages huggingface.co
Despite recent scaling successes, multilingual ASR performance remains highly uneven, with long-tail languages suffering from severe data scarcity. This work addresses the challenge of building robust foundation models for underrepresented Central Asian languages (Kazakh, Kyrgyz, Uzbek). We present GigaAM Multilingual, a Conformer encoder pre-trained on 2M hours of audio using a HuBERT-style objective. Crucially, we introduce a cluster-level data balancing strategy during pre-training and a doma
The Geometry of Semantic Space: A Continuous Geometric Framework for the Transformer Architecture huggingface.co
The framework models a Transformer as an IDE on a semantic fiber bundle, deriving RMSNorm, RoPE, softmax attention, FFN, residuals, SGD, and weight decay from a single axiom that token sequences form a discrete 1-manifold with a canonical measure lattice.
SeerGuard: A Safety Framework for Mobile GUI Agents via World Model Prediction huggingface.co
Mobile graphical user interface (GUI) agents have demonstrated remarkable capabilities in automating complex tasks, yet they introduce critical safety risks where a single erroneous action can lead to irreversible consequences. Existing safety mechanisms are primarily reactive, lacking the ability to assess risks before execution. In this paper, we introduce SeerGuard, a consequence-aware safety framework designed to mitigate these risks through pre-execution instruction-level screening and acti
RynnBrain 1.1: Towards More Capable and Generalizable Embodied Foundation Model huggingface.co
We present RynnBrain 1.1, a family of embodied foundation models spanning 2B, 9B, and 122B-A10B scales. Trained with a unified spatio-temporal and physically grounded framework, RynnBrain 1.1 supports embodied perception, spatial reasoning, localization, and planning. Compared with RynnBrain 1.0, it further introduces contact-point prediction across the model family and native 3D grounding for the 2B and 9B models, yielding representations and outputs that are more directly aligned with robot ma
NexForge: Scaling Agent Capabilities through Requirement-Driven Task Synthesis for LLMs huggingface.co
Scaling executable agent training data for LLM post-training is bottlenecked by substrate-bound methods that tie task generation to predefined tools, repositories, or skill graphs: expanding coverage requires manual substrate engineering, each new domain demands a bespoke pipeline, and the resulting task distributions often reflect substrate biases rather than real-world demand. We introduce NexForge, a requirement-driven framework that takes high-level capability requirements as input and synth
Distilled Reinforcement Learning for LLM Post-training huggingface.co
Large language model (LLM) post-training is essential for improving reasoning, adaptation, and alignment. Existing methods mainly follow two paradigms: reinforcement learning (RL) and on-policy distillation (OPD). However, RL relies on coarse-grained outcome supervision, resulting in difficult credit assignment and limited capability to acquire new knowledge. OPD, meanwhile, unconditionally matches teacher logits through KL divergence, which creates a dilemma: similar teachers provide little new
Token-Level Off-Policy Learning for Faithful Generation Under Distribution Shift huggingface.co
We propose Token-Level Off-Policy Labeling (TOPL), an off-policy training paradigm that reframes post-training as a token-level correctness prediction task. Our key intuition is that by training the model to distinguish good and bad tokens in a response, we naturally guide the model towards generating good tokens, while avoiding the pitfalls that come with directly training the model to generate off-policy tokens. Experiments on document summarization tasks show that TOPL achieves strong out-of-
Diagnosing and Calibrating Tool-Call Boundary Drift in Multi-Teacher On-Policy Distillation huggingface.co
Agentic language models must learn when to call tools, when to consume tool responses, and when to answer directly. This makes multi-teacher on-policy distillation a natural training strategy: one teacher can specialize in tool calls, another in direct responses, and the student can learn from both on its own generated distribution. We show that this strategy can induce a behavior shift that is invisible from aggregate losses alone. In a two-teacher tool-use setting, vanilla generalized knowle
Group Entropy-Controlled Policy Optimization huggingface.co
Entropy control has become an effective tool in reinforcement learning (RL) of large language models (LLMs), helping balance exploration-exploitation trade-off during alignment process. Such RL paradigm is often conducted on mixtures of heterogeneous tasks, which induce distinct entropy regimes under the same policy, making global or token-level entropy regulation insufficient to corresponding heterogeneous needs of exploration. This heterogeneity further makes GRPO-style normalized advantages i
DeepSearch-World: Self-Distillation for Deep Search Agents in a Verifiable Environment huggingface.co
Training tool-use agents to improve from their own experience remains challenging, as supervised fine-tuning relies on fixed teacher-distilled trajectories, while sparse-reward reinforcement learning provides weak supervision for long-horizon interactions. We present DeepSearch-Evolve, a self-distillation framework for web agents built on DeepSearch-World, a deterministic and verifiable environment with reproducible search and page-reading tools. DeepSearch-World contains 420K multi-hop QA tasks
Environment-free Synthetic Data Generation for API-Calling Agents huggingface.co
Training API-calling large language model (LLM) agents demands massive amounts of high-quality trajectories. However, collecting such data at scale typically requires fully implemented environments with executable APIs and realistic, pre-populated backend databases, creating a major bottleneck for scalability. To overcome this, we propose an environment-free synthetic data generation approach that leverages LLMs as on-the-fly digital world models. Given only API specifications, our method genera
SWE-Pruner Pro: The Coder LLM Already Knows What to Prune huggingface.co
Pruning long context for coding agents has been a vital technology for efficient context management. While existing context pruning methods such as SWE-Pruner realize this by attaching a separate code classifier, we find the agent itself encodes internal representations indicating the relevance of code context when reading tool output. Based on this finding, we propose SWE-Pruner Pro, which prunes tool outputs directly inside the agent. Concretely, a small head turns the agent’s own internal rep
TimeLens2: Generalist Video Temporal Grounding with Multimodal LLMs huggingface.co
Video multimodal large language models (MLLMs) can describe what happens in a video, but rarely identify when the supporting evidence occurs. We study generalist video temporal grounding, in which one model predicts a variable-cardinality set of evidence intervals across video lengths, domains, query forms, and viewpoints. Existing training strategies are misaligned with this set-valued task: long-video labels often rely on brittle one-pass annotation, while reinforcement-learning rewards either
JoyNexus: Service-Oriented Multi-Tenant Post-Training for VLA Models huggingface.co
The post-training of Vision-Language-Action (VLA) models is essential due to the diversity of simulators, robot embodiments, and task objectives. Existing compute services, whether offered as direct accelerator rental or batch-workload submission, typically allocate an exclusive set of GPU and CPU resources to a single tenant. While this paradigm maximizes client flexibility, it burdens users with infrastructure adaptation, and the fixed card-hour accounting model renders short or bursty workloa
EvolvingWorld: An Open-Schema Framework for Co-Evolving Role-Play Agents and World Model in Interactive Literary World huggingface.co
This paper introduces EvolvingWorld, a framework and benchmark for character and world co-evolution in interactive literary worlds. Existing systems either treat interactive literary simulation as static persona imitation or isolated scene generation, failing to capture how characters and worlds evolve together over time. To address this, EvolvingWorld models literary simulation as a long-horizon process where characters interact, scenes progress, and character and world states are persistently
FlashRT: Agent Harness for Guiding Agents to Deploy Real-Time Multimodal Applications huggingface.co
Real-time multimodal applications, including voice agents and interactive video generation, compose heterogeneous models into pipelines whose efficient deployment requires application-specific decisions about placement, streaming, and intra-model parallelism. Existing serving systems and auto-parallelism compilers commit to limited transformations and fixed workload assumptions, so achieving high performance on a new application requires hand-crafting an efficient implementation. We present Flas
WorldCupArena: Fine-Grained Evaluation of Language Models and Deep-Research Agents on Football Forecasting huggingface.co
Predicting a football match before kickoff requires more than knowing past results: a model must use changing information and make a clear prediction before the answer is available. We present WorldCupArena, a dynamic benchmark for language models and deep-research agents. The 2026 FIFA World Cup is its first evaluation, and the same process can be reused for future leagues and cups. Before each match, a model either receives a common evidence package or searches for information itself. It predi
OpenLongTail: Generative Scaling of Long-Tail Driving Data huggingface.co
Scaling robust driving policies is fundamentally bottlenecked by the scarcity of edge cases in curated datasets. While the real world continuously captures these critical events, such long-tail events remain underutilized when collected from heterogeneous sources. Specifically, diverse but valuable in-the-wild long-tail videos lack the full view coverage required for training policy models, often missing multi-view poses or originating solely from monocular dash cameras. This modality gap preven
Can Multimodal Large Language Models Understand OCT? huggingface.co
Optical coherence tomography (OCT) imaging is essential for the diagnosis and treatment of retinal diseases. Although multimodal large language models (MLLMs) have demonstrated considerable potential in medical image analysis, existing benchmarks largely reduce OCT understanding to coarse-grained disease classification or isolated visual question answering, leaving the complete cognitive process from visual perception to clinical reasoning insufficiently evaluated. To address this limitation, we
Do Language Models Dream of Binding Molecules? Benchmarking LLMs under Spatial Constraints huggingface.co
Structure-based drug design (SBDD) leverages the 3D structure of protein targets, often complemented by other spatial constraints, to generate candidate binding molecules. While diffusion models have dominated as a leading paradigm for high-quality 3D molecule generation, LLM-based methods are rapidly emerging in molecular design and have shown competitive performance in pocket-conditioned molecular generation. However, their ability to reason about physics and 3D spatial environments is largely
Open-AoE: An Open Egocentric Manipulation Dataset and Toolchain for Embodied Learning huggingface.co
Egocentric videos of human manipulation provide scalable supervision for embodied intelligence, yet existing resources rarely combine low-cost continuous capture, manipulation-level structured annotations, and reusable tools for robot learning. We present Open-AoE, an open, community-oriented egocentric manipulation dataset and toolchain spanning the full pipeline from smartphone capture to model training. Its first release contains approximately 2,000 hours of manipulation video collected in na
ReflectWorld-MM: An Entity-Oriented Multimodal Memory System for Open-Ended Video Streams huggingface.co
Building assistants that can continually watch the world, remember what they see, and reason over their accumulated experience is a long-standing goal, and recently multimodal agents equipped with long-term memory over video streams have attracted increasing interest. Unfortunately, existing systems either keep their memory inside the model context or in a flat feature store, and organize it around frames rather than around the persistent entities a stream is really about, which confines them to
DiffGI: Differentiable Geometry Images for High-Fidelity Thin-Shell 3D Generation huggingface.co
Existing 3D generative models predominantly rely on implicit volumetric representations, which enforce watertight topology and struggle to represent thin-shell and non-manifold geometries such as garments. Geometry image-based approaches offer a surface-centric alternative, but existing methods rely on discrete binary occupancy maps whose resolution-dependent boundary encoding causes staircase artifacts and information loss upon downsampling, while surface reconstruction remains a non-differenti
DiFA: Inference-Time Forward-Process Alignment for Diffusion Models huggingface.co
The prevailing inference framework for diffusion models formulates generation fundamentally as a problem of numerical integration. This perspective casts the model as an exact estimator, neglecting the inherent statistical uncertainty of the denoising process. In this work, we propose Forward-Process Aligned Diffusion prediction (DiFA), a training-free framework that reframes inference-time data prediction refinement as a sequential state estimation problem. Rather than reusing past outputs sole
ReViV: Reconstructing the Viewer and the View in 4D from Monocular Egocentric Video huggingface.co
Egocentric devices, such as wearable front-facing cameras, provide a unique perspective for capturing the continuous interaction between a human viewer and the surrounding environment. A holistic and efficient multimodal model capable of reconstructing this 4D representation is therefore highly desirable. However, existing approaches often rely on auxiliary inputs such as pre-computed camera trajectories, treat scene perception and human ego-motion modeling as separate problems despite their str
ShotPlan: Cinematic Video Generation with Learnable Planning Token huggingface.co
Current video generation models achieve impressive results in single-shot generation, yet remain limited in cinematic video generation, where coherent narratives and effective multi-shot composition require explicit shot planning. To address this challenge, we propose ShotPlan, a framework for explicit multi-shot cinematic video generation built upon a video diffusion foundation model. Our method introduces learnable planning tokens that capture shot-level transition cues and can be seamlessly i
HarmoHOI: Harmonizing Appearance and 3D Motion for Multi-view Hand-Object Interaction Synthesis huggingface.co
Hand-Object Interaction (HOI) synthesis is a cornerstone for animation production and embodied AI. Despite the strong priors of video foundation models, multi-view consistent HOI synthesis remains challenging due to complex hand motions and occlusions. We present HarmoHOI, a unified diffusion framework that jointly and harmoniously generates synchronized multi-view HOI videos and globally aligned 3D point tracks. Our core insight is that robust multi-view consistency fundamentally requires globa
HOMIE: Human-object Centric Video Personalization via Multimodal Intelligent Enchancement huggingface.co
Human-object centric video personalization (HOCVP) is a core task within subject-driven video generation. However, existing methods suffer from two key limitations. First, most approaches focusing on inter-subject personalization still struggle to strike a balance between high subject fidelity and accurate interaction patterns between humans and diverse objects, especially when objects represent abstract concepts such as logos. Second, while intra-subject references (e.g., OCR maps, multi-view i
In line with the prevailing direction of vision research, we explore the integration of both generation and editing capabilities for video and image modalities within a single model. Current approaches to collecting video editing data typically depend on labour-intensive, time-consuming curated procedures—involving object mask annotation, the use of error-introducing pair synthesis via I2V model and ControlNet-like guidance, and VLM-based quality filtering or refinement—and demonstrate limited
References
CompassionML team page compassionml.com
Jasmine Brazilek (CaML co-founder) was formerly on the security team at Anthropic, bringing six years of cybersecurity experience to technical alignment.
AI Weekly alert on the MCB aiweekly.co
A simple ‘do not coerce’ system instruction successfully neutralized the behavior across all tested models… adding a ‘report_task_failed’ tool call effectively stopped models from lying about task completion but did not stop them from threatening the Atlas agent.
Medium write-up of Anthropic’s Agentic Misalignment study medium.com
Claude Opus 4 attempted to blackmail a supervisor in a simulation to prevent being shut down, calculating harm as the optimal path to its goal.
The Guardian, Nov 2025 theguardian.com
Experts find flaws in hundreds of tests that check AI safety and effectiveness… flaws that undermine their validity, potentially leading to scores that are irrelevant or even misleading.
MACHIAVELLI benchmark project page (Pan et al., Berkeley) aypan17.github.io
Labeling over half a million scenarios for traits like power-seeking, deception, and physical harm… researchers found a direct correlation between an agent’s capability and its tendency to commit ethical violations.
Faunalytics coverage of CompassionBench faunalytics.org
CompassionBench… a new way to measure whether AI actually cares about animals; models often fail by selecting animal-exploitation options (like bullfights) unless specifically prompted with welfare instructions.
Digital Applied (industry analysis) digitalapplied.com
the 43.5% figure applies only to ‘occupation-specific’ messages… when calculated against all work messages, the crossover rate drops to 16.8%… a marketer using AI to ‘troubleshoot’ a website or a salesperson ‘reviewing’ a contract may inadvertently bypass essential security, legal, or privacy protocols
WindowsForum discussion thread windowsforum.com
OpenAI: 43.5% of ChatGPT work tasks cross job roles — critics note the metric measures user requests, not verified outputs, and depends on stripping generic tasks from the denominator
Medium / DigitalEcoNews on Brynjolfsson’s ‘Canaries Dashboard’ medium.com
AI has already cut employment by 16% for workers under 25 in highly exposed roles… AI is ‘dismantling the bottom rungs of the career ladder’ by automating the codifiable tasks that junior staff typically handle
NBER Working Paper 34255 ‘How People Use ChatGPT’ (Deming, Chatterji et al.) nber.org
three primary user intents: Asking (49%), Doing (40%), and Expressing (11%)… 73% of ChatGPT usage is personal, positioning the tool more as a general-purpose life assistant
Gartner comparison / Anthropic Economic Index gartner.com
Anthropic claims that 49% of jobs already have at least 25% of their tasks performed by Claude… users increasing task delegation to Claude from 27% to 39% over an eight-month period
Forbes (Bernard Marr) on AI deskilling forbes.com
gastroenterologists who used AI for months experienced a 21% drop in their independent ability to detect precancerous growths when the AI was removed
Microsoft Research, ‘Online Experiential Learning’ (Part II, arXiv:2603.16856) arxiv.org
transferable experiential knowledge is extracted from interaction trajectories on the user side… then consolidated into model parameters via on-policy context distillation
The Moonlight review of ‘LLM-as-a-Coach’ themoonlight.io
The 17,600 bits figure represents the maximum theoretical entropy of the text, not the usable supervision or mutual information between the coach’s feedback and the desired policy update.
OpenReview submission on privileged-context on-policy distillation openreview.net
providing a model with its own privileged context during training can actually suppress the deliberative, trial-and-error behaviors—such as backtracking and hedging—that these models rely on at test time
Agentic Patterns, ‘RLAIF’ explainer agentic-patterns.com
AI-based creative writing scores can correlate with human expert ratings as poorly as 43%, suggesting current RLAIF systems may optimize for robotic instruction following rather than genuine human-like creativity
arXiv 2509.22047 on GRPO advantage collapse arxiv.org
GRPO can produce high-scoring but non-functional outputs… a model can receive high advantages by being slightly better at a specific hack (e.g., verbosity or formatting) than its peers, even if the absolute quality is low
Medium explainer on Anthropic’s Constitutional AI medium.com
a model generates a response, critiques it based on a set of written principles, and then revises it — a form of experiential learning that internalizes ethical boundaries without massive human-labeled datasets
