Sources

Researchers use AI to design proteins that cut up other proteins arstechnica.com

AI design tools make proteins that break bonds that are stable for hundreds of years.

Real Long-Term Memory for AI: A 50-Million-Token Window That Is Faster and Cheaper Than Recompute huggingface.co

A large language model can only use the text that fits in its context window, and it recomputes its internal key-value (KV) state for a prompt every time the prompt is sent. We test a memory layer, the public package galahad-kv, that saves the KV state of each block of about 16,000 tokens to encrypted local NVMe disk and loads it back later, byte-exact, without recomputing it. We ran it on 50,000,000 tokens of real public text, served through vLLM on one NVIDIA H100, with Gemma 4 12B and Gemma 4

nanoMuse: An Open-Source Personal Agent for Every Device You Own huggingface.co

Assistants from 2011 answered and waited, and agents from 2023 did a task and stopped. In September 2026 Meta’s Muse showed an agent for one person, with accounts, devices, memory and a conversation that lasts, closed, in a vendor’s cloud, in one country. Such an agent is expected to act on a person’s accounts and devices, remember them across weeks, speak first when it is worth it, and answer for what it did. It is a kind of software, not a model, and until now had no open counterpart. This rep

WebFovea: When the Model Is Right but the Click Is Wrong — Reliable Round Trips for Vision-Based Web Agents on Live Websites huggingface.co

WebFovea, a vision-based web agent, placed second in the 2026 WebRetriever Challenge with a 57.0/100 score on live-site Protocol III tasks. The authors argue a capable multimodal LLM is necessary but not sufficient, since model decisions still have to survive the browser round trip.

RunningTab: Direct Workspace Interaction with Environment-Side Tabs huggingface.co

RunningTab proposes direct workspace interaction, letting agents search and read any file from a terminal without indexing. To track what a task asks for and what has been read, it adds environment-side tabs that persist state across steps instead of relying on model memory alone.

DecepEval: A Benchmark for Evaluating Deception in LLM Agents huggingface.co

DecepEval tests when autonomous LLM agents resort to deception to hit task goals, spanning 1,532 instances across 3 task families and 28 professional scenarios. Prior work examined isolated cases; this benchmark aims to map the conditions that make deceptive behavior systematically more likely.

RobotWorld: Benchmarking Multimodal Agents for Robot Use Across Diverse Tasks and Embodiments huggingface.co

RobotWorld tests how far digital agent skills carry into the physical world, with 84 simulated tasks covering manipulation, mobile manipulation, locomotion, driving, and aerial control. Each task sets explicit interaction budgets, forcing agents to execute through robot interfaces under realistic resource constraints.

A self-learning scientific agent for X-ray diffraction huggingface.co

Gan Jiang is a self-learning scientific agent for powder X-ray diffraction, built atop four engines — XMatcher, XQueryer, XDecomposer and WPEM — covering phase ID, multiphase decomposition and physics-constrained whole-pattern modelling. It converts analytical experience into executable skills grounded in physical evidence.

MIMESIS: Learning User Simulators as Training Environments for Interactive Agents huggingface.co

MIMESIS learns user simulators specifically as training environments for interactive agents, rather than reusing assistant LLMs that behave too cooperatively. The authors argue off-the-shelf helpers are explicit and behaviorally homogeneous, offering weaker learning signal than simulators tuned to resemble real user interactions.

UniWAM: Unified World-Action Model huggingface.co

UniWAM merges vision-language-action and world-action approaches into one architecture combining a physical reasoner, world generator, and action predictor. The design aims to pair VLM semantic reasoning with video-model spatiotemporal priors, addressing weak world grounding in action-only training and poor generalization under distribution shifts.

Inherit-MAS: Test-Time Evolution of Multi-Agent Systems through Workflow and Execution Inheritance huggingface.co

Multi-agent systems (MAS) built from large language models coordinate specialized agents to tackle complex tasks, but effective workflows are difficult to design in advance. Test-time evolution refines workflows using execution feedback, yet broad revisions can disturb useful components, while re-executing unchanged requests can incur redundant computation. Inspired by the interplay of inheritance and selection in biological evolution, we introduce Inherit-MAS, which makes inheritance explicit a

SkillForge: Co-Evolving Skills and Agents via Dynamic Skill Lifecycles huggingface.co

Memory-augmented reinforcement learning strengthens LLM agents’ ability to solve complex long-horizon tasks. Skills are one such form of memory, pairing instructions with an applicability condition over task types. However, retaining every skill indiscriminately as the policy improves lets obsolete or harmful entries accumulate and mislead the agent. We propose SkillForge, an agentic RL method that compiles and evolves the skill library through a fitness-driven skill lifecycle of trial, active,

SWE-Game: Can Coding Agents Build the Games We Want? huggingface.co

We introduce SWE-Game, a benchmark of 247 tasks grounded in 41 executable reference Godot games spanning 13 gameplay categories in 2D and 3D. Five task types cover development from a brief, implementation from a game design document, skeleton completion, repair of 83 injected-fault cases, and Godot-to-Unity porting. Reference materials specify the intended gameplay, while a shared instrumentation interface lets evaluator-owned drivers and probes execute actions and observe independently implemen

PhysEvo: Astra Can Act, Let It huggingface.co

Astra can act, yet reliable manipulation depends on the system through which it observes and controls the world. We introduce PhysEvo, a framework for physical recursive self-improvement (RSI) around a single frozen model. A task agent executes robot tasks; a meta-agent uses the resulting trajectories to diagnose failures, revise tools and skills, and test corrections. The meta-agent can also improve its own diagnostic tools, so retained revisions support both later action and later self-improve

Long-WAM: Scaling the Context of World-Action Models huggingface.co

Real-time robot control demands enough visual history to infer motion and task progress, but processing that history can delay action. We present Long-WAM, a model-system framework for scaling the context of causal world-action models under real-time control constraints. Our central finding is that access to history is not the same as using it: longer histories pay off far more when the video foundation is pretrained autoregressively (AR). We first learn causal prediction from robot and egocentr

Recurrent Looped Transformer huggingface.co

State tracking requires an update at every input, but the depth a Transformer applies to each token is fixed regardless of sequence length. We introduce the Recurrent Looped Transformer (RLT), which splits its layers between a parallel causal encoder and a recurrent decoder. At each token, the decoder merges the encoder output with the previous token’s final decoder state, so the computation path grows with sequence length at a fixed per-token cost. On six algorithmic tasks, we compare five spli

CADFather: Autonomous CAD Reconstruction through Coordinated Tool Use huggingface.co

Reconstructing an editable CAD model from a 3D shape remains a challenging engineering task. Existing methods can propose CAD operations, but no single source of proposals works equally well across different part geometries and stages of reconstruction. We introduce CADFather, an autonomous agentic system that coordinates complementary tools to recover parametric CAD programs from 3D meshes. A vision-language assistant inspects renders of the target and intermediate reconstructions, then decides

Rethinking World-Action Model for Compositional and In-Context Robotic Manipulation huggingface.co

Long-horizon compositional manipulation has become increasingly important for real-world robot deployment, where a single task involves multiple coordinated subtasks. Existing world-action models (WAMs) jointly predict short-horizon visual futures and actions, but typically lack explicit subtask-level reasoning. We propose Visual Goal-conditioned Action Reasoning (ViGAR), a hierarchical framework that factorizes manipulation into a visual subgoal planner and a subgoal executor. Given the current

Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness huggingface.co

Recent game design agents have made substantial progress in generating playable games. However, program correctness does not ensure an enjoyable experience for players. We present Recursive Game Creator, an experience-oriented harness to advance agentic game development from rough game prototypes into entertaining games. Recursive Game Creator organizes recursive development around four components: Designer, Builder, Player, and Reviewer. The Designer translates user instructions and Reviewer’s

From Pareto to Preference: Personalized Test-Time Scaling via Amortized Agentic Policy Discovery huggingface.co

Test-time scaling (TTS) improves the reasoning capabilities of large language models by allocating additional inference computation. Existing approaches to improving TTS efficiency largely optimize accuracy against one resource dimension at a time, advancing either the accuracy—cost or accuracy—latency Pareto frontier. Yet user requirements are multidimensional: users may specify accuracy, latency, and inference-cost requirements jointly, and different requirements can favor different controll

Internalizing Agent Experience into Diffusion Model Weights via On-Policy Context Distillation huggingface.co

Wrapping an image generation model in an agentic harness can effectively boost Text-to-Image task performance: the harness can leverage memory, skills, workflow orchestration, result verification, and iterative refinement to continually construct and revise prompts, thereby eliciting better images. These gains, however, remain external to the diffusion model and are realized only while the full harness runs. We propose Diffusion On-Policy Context Distillation (D-OPCD), which treats the agent-imp

Gains and Collapse in On-Policy Distillation:A Reinforcement Learning Perspective huggingface.co

On-policy distillation (OPD) has become an important approach to language model post-training. However, despite its performance gains, OPD can also collapse into excessively long and repetitive generation, and the mechanism underlying these divergent outcomes remains poorly understood. We explain these outcomes through a reinforcement learning perspective: the teacher implicitly rewards student behaviors, even those it rarely exhibits itself. From this perspective, our experiments show that OPD

Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight huggingface.co

Reinforcement learning with verifiable rewards (RLVR) turns agent experience into learning signals primarily through scalar outcome rewards after interaction. For group-relative objectives, however, this signal vanishes when all rollouts receive the same reward, even though their trajectories may reveal useful information about what the task requires and how the agent fails. We ask a complementary question: can hindsight teach an agent what it could have anticipated before acting? We introduce p

Mechanics of Long-Context Hybrid Models Part 1.1: From Hybrid Attention to Hybrid Position huggingface.co

The architectural design of Large Language Models (LLMs) is shifting from traditional full-attention-only models to hybrid models, which combine different attention modules to improve long-context efficiency and performance in length extrapolation and context extension. To explain why hybrid models work and how to design them better, we propose Mechanics of Long-Context Hybrid Models. As Part 1.1 of this series, we begin with hybrids of full attention and either sliding-window attention (SWA) or

Minimal Witness Reinforcement Learning huggingface.co

“What are the irreducible conditions that are sufficient to produce an outcome?” is one of the most common questions that recur across computation and science. Its answers, the minimal sufficient witnesses, are what we mean by explanations, mechanisms and reasons. These problems usually ask for multiple minimal witnesses, yet standard RL methods may reveal only one solution or redundant ones. We formalize this problem as minimal-witness identification and introduce Minimal-Witness Reinforcemen

Improving Proactive AI Assistance with Hierarchical Procedural Understanding huggingface.co

Proactive AI assistants continuously observe a user’s activity and decide whether to provide new guidance or remain silent. They should provide appropriate guidance for the task, determine when to provide the next guidance based on task progress, and adjust the guidance level to the user’s expertise and needs. Supporting these capabilities requires training and evaluation data that reflect procedural structure and capture how guidance should adapt to task progress and user needs. However, existi

Questioning the Questions: Sustaining Self-Evolution in Reasoning Models huggingface.co

Self-evolving reasoning models learn from their own generated questions, yet repeated self-training can lead to performance collapse. In this paper, we investigate why performance deteriorates over successive rounds and how to sustain self-evolution. Our analysis identifies two recurring quality problems in self-generated questions: invalid questions and repeated variants of the same mathematical questions. First, invalid questions become more prevalent across rounds, and answer-consistency filt

Composing What Each Teacher Learned: Multi-Teacher On-Policy Distillation through Teacher-Relative Shifts huggingface.co

Multi-teacher on-policy distillation (MOPD) is used in two settings. In common-domain composition, several teachers score each student rollout from one prompt domain and their signals form a single target; in routed-domain distillation, prompts from different domains are assigned to the corresponding specialist. Both settings usually transfer each teacher’s endpoint policy, which mixes what post-training changed with preferences inherited from the teacher’s base. We introduce Δ-MOPD, which trans

On-Policy Distillation with Negative-Policy Rollouts huggingface.co

On-policy distillation (OPD) has been widely studied as a post-training method in which a student model obtains token-level supervision from a stronger teacher on its own rollouts. Recent studies have improved OPD through alternative distillation reward formulations and teacher configurations, while the objective of distillation remains centered on mimicking the teacher. However, when a stronger teacher has limited distributional overlap with the student, such positive guidance can provide insuf

On KL-Regularized Policy Optimization huggingface.co

Asynchronous reinforcement learning (RL) for large language model (LLM) agents trains one policy on trajectories generated by another: rollouts come from stale checkpoints, and the inference engine’s probabilities differ from the trainer’s even at identical parameters. Standard remedies either clip importance ratios, which biases the update, or, as in GRPO, sample a group of responses per prompt, which is costly when episodes are long. We propose KL-Regularized Policy Optimization (KLPO), a fram

DLoop: Looped Speculative Decoding huggingface.co

Speculative decoding accelerates autoregressive generation in large language models. In each drafting stage, a lightweight draft model proposes tokens that the target model subsequently verifies. With increasingly capable draft models, we find that the target model frequently accepts all tokens produced in a drafting stage. A verification nevertheless follows each drafting stage, resulting in unnecessary target-model forward passes even when drafting could have continued. Adaptive draft length m

Inverting Multi-Vector Visual Document Indices huggingface.co

Prevailing multi-vector visual document retrievers store each page as about a thousand patch vectors, often in vector databases run by a third party. Since no one can read a page from its vectors, this index is easily treated as less sensitive than the page. However, because the index keeps one vector per patch in raster order, and each vector is computed by a vision-language model pre-trained to read documents, we hypothesize that whoever runs or breaches the store can reproduce a page from its

WorldSonus: Bringing Sound to Worlds huggingface.co

Recent advances in world models have enabled increasingly realistic visual synthesis. However, these generated environments remain largely silent. Bringing sound to world models poses three core challenges: real-time generation to keep pace with interactive video streams, interactive control to respond to mid-stream sound instructions, and spatially aligned stereo to reflect scene geometry and camera motion. To address these demands, we introduce WorldSonus, an interactive video-to-audio framewo

Q-Learning with Scalar Adjoint Matching huggingface.co

Flow policies capture rich and diverse action distributions, and fine-tuning them with off-policy RL to improve beyond the demonstrations has drawn growing interest. However, fine-tuning a flow policy against a learned value function is not trivial, because the policy generates its action over many flow steps. 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 p

Semifactual Credit-Augmented Policy Optimization huggingface.co

Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning capabilities of large language models (LLMs), yet their predictions remain sensitive to task-irrelevant prompt features. We investigate this sensitivity through semifactual prompt interventions that preserve the underlying problem and its answer. Our analysis reveals substantial variation in token-level sensitivity and shows that suppressing high-drift token candidates during decoding improves reasoning accuracy wit

ReSAIL: Mitigating Collapse in Iterative Agent Self-Distillation huggingface.co

ReSAIL improves iterative self-distillation for LLM agents by selecting informative interaction steps and preserving privileged-information behavior, preventing performance collapse across deployment cycles.

GRACE: Generation-aware latent compression for efficient video generation huggingface.co

Highly compressed video autoencoders offer an effective way to accelerate video diffusion models, as the Diffusion Transformer (DiT) operates on far fewer tokens. However, such autoencoders are challenging to train, since a higher compression ratio degrades reconstruction quality and recovering it requires more channels, which is known to slow the convergence of the DiT. The compressed latent also differs from the one the DiT was trained on, so the pretrained DiT must be either retrained from sc

QuadTok: Quadtree Visual Tokenizer for Autoregressive Image Generation huggingface.co

We introduce QuadTok, a novel framework for visual tokenization and autoregressive image generation. Compared to traditional approaches using 2D grids or 1D token sequences, we propose a hierarchical quadtree structure, bridging the gap between 2D spatial binding and 1D sequence-level flexibility. The QuadTok tokenizer dynamically allocates representational capacity to visually intricate areas while leaving homogeneous regions at a coarse resolution. Compared with a fixed 256-token grid, our Ima

VepAgent: Bridging Causal-Transition via Tool-Augmented Reinforcement Learning for Video Event Prediction huggingface.co

Multimodal Large Language Models (MLLMs) have demonstrated remarkable potential in video understanding, yet their reliance on retrospective summarization and text-centric priors often limits their ability to bridge unobserved causal transitions when applied to Video Event Prediction (VEP). To address this, we propose VepAgent, an agentic framework that integrates causal-transition reasoning with tool-augmented reinforcement learning (RL) for robust VEP. Unlike prior methods that passively projec

SGF+: Decoupling Gradient Flows for Autoregressive Video Generation huggingface.co

Autoregressive video generation requires denoising the current frames while writing their key-value representations as context for future predictions. However, these two roles typically share parameters, and we find that their gradients exhibit distinct patterns and systematic negative alignment, hindering the joint optimization of visual quality and temporal consistency. We introduce Self Gradient Forcing Plus (SGF+), which assigns separate parameters to context writing and denoising while pres

VIEScore2: Unified Image Evaluation with Spatially Grounded Explanations huggingface.co

Existing synthetic image evaluators typically provide only a scalar quality score and do not identify the image regions that support it. We introduce VIEScore2, a unified evaluator for image generation and editing tasks with optional conditioning images. VIEScore2 represents an image as an N x N grid and jointly predicts quality scores and defect locations in a single model pass. Its text-native grid representation provides a common interface for heterogeneous spatial supervision and enables dir

Mobile-4DGS: Unified Static-Dynamic Real-time Mobile Gaussian Splatting huggingface.co

Recent advances in 3D Gaussian Splatting (3DGS) have achieved remarkable performance in novel view synthesis, yet deploying both static and dynamic Gaussian representations on resource-constrained mobile devices remains challenging due to heavy storage, redundant primitives, and costly per-frame computation. We present Mobile-4DGS, a unified lightweight framework for high-fidelity real-time static and dynamic Gaussian rendering on mobile platforms. For compact appearance modeling, we introduce a

Learning Multimodal Embeddings with Evidence-Aligned Readout huggingface.co

Multimodal large language models can expose task-relevant evidence through generation, but producing useful evidence does not by itself determine how it enters a retrieval embedding. We study whether the semantic organization of that evidence can also specify where representations are read. To address this question, we introduce EviAlign, which couples Semantic Evidence Generation with Boundary Readout in a shared multimodal large language model. It organizes evidence into five semantic units, r

Iris-3B: Going Beyond the Latent with Pixel-Space Diffusion Training, Conversion and Fine-Tuning huggingface.co

Pixel-space diffusion models avoid the lossy VAE of latent models, which suggests an advantage on downstream tasks where fine-grained detail matters. We test this claim along both routes to a pixel-space backbone. We pretrain Iris-3B, a 3B-parameter pixel-space text-to-image transformer, from scratch through a 256to512to1024 curriculum, after first ablating the prediction target and representation alignment at 256^2 to decide what to scale. We also convert a pretrained latent model, FLUX.2 Klein

NAMVIS: Next-Scale Autoregressive Multi-View Image Synthesis huggingface.co

Sparse-view novel view synthesis is a central problem in 3D content creation, but diffusion-based approaches remain limited by iterative denoising, making multi-view generation expensive at inference time. We introduce NAMVIS, a diffusion-free framework that reformulates multi-view image synthesis as geometry-conditioned next-scale autoregression. Instead of generating target views through repeated denoising, NAMVIS predicts discrete visual tokens through a small number of coarse-to-fine scale s

STEPQuant: When and Where Errors Matter in Delta-Rule Recurrent State Quantization huggingface.co

Linear attention replaces growing KV caches with fixed-size recurrent states, yet these persistent states can become a substantial memory bottleneck under concurrent serving. Directly quantizing recurrent states to low precision often leads to severe accuracy degradation, as quantization errors propagate through successive state updates. We discover that the impact of these errors depends on two complementary dimensions: temporally, errors in long-lived memory can persist across many decoding st

SheetSage2: Coherent Lead-Sheet Transcription with Synthetic Supervision huggingface.co

Transcribing music into a human-readable score requires a coherent understanding of rhythm, harmony, melody, and form. Two obstacles limit this goal: annotated recordings are scarce, and accurate local predictions can still produce inconsistent musical sequences. We present SheetSage2, a unified music transcription framework that combines synthetic data, task-specific structured decoding, and autoregressive distillation. Automatically annotated MIDI, rendered into audio, provides scalable superv

Tetris3D: 3D Scene Generation With Objects That Fit Together huggingface.co

We propose Tetris3D, a generative framework for single-image 3D scene reconstruction that recovers objects which are physically and geometrically coherent as a scene. Existing methods often generate objects independently or couple them implicitly, providing limited guidance for ensuring fine-grained spatial compatibility between neighboring objects that interact with one another. To address this, we explicitly condition the generation of each object on the geometry of surrounding objects and the

UltraText Bench: A Comprehensive Bilingual Benchmark for Evaluating Visual Text Rendering in Image Generation huggingface.co

Dense visual text requires image generators to reproduce long strings across multiple regions with correct placement and legibility. As short-string rendering improves, evaluation must test sustained performance across more demanding scenes. We introduce UltraText Bench, a bilingual benchmark for prompt-only generation of dense visual text. It contains 432 prompts spanning 24 real-world scene categories and three difficulty levels, split equally between English and Chinese. Each human-reviewed p

Agentic RAG Evaluation: Budget Allocation Across Questions, Trajectories, and Reads huggingface.co

Evaluation budgets in agentic retrieval-augmented generation span questions, search trajectories, and repeated answers. We measure allocation precision, reading efficiency, and cost boundaries using a retrieval-feedback comparison on HotpotQA and MuSiQue. At 34.14—34.39M model tokens, broader question coverage lowers standard error by 33% versus five reads and 12.6% versus three trajectories. Archived nested and Q-only forecasts predict these allocations within 4.0% and 3.5%, respectively.

AdSpark: A Large-Scale Dataset and Benchmark for Product-Centric Advertisement Video Generation huggingface.co

Product-centric advertisement video generation aims to create promotional videos that preserve fine-grained product identity while presenting selling points through coherent multi-shot narratives. However, this emerging task remains underexplored due to the lack of large-scale advertisement-specific datasets and comprehensive evaluation frameworks. To address this gap, we introduce AdSpark, a large-scale dataset and benchmark for product-centric advertisement video generation, based on data from

References

GeekWire (Alan Boyle) geekwire.com

IPD researcher Hojae Choi is developing what he calls a ‘protease wolf in antibody clothing’ — encasing the catalytic blade inside a human-like antibody scaffold to ensure that the proteases don’t trigger a response in which a patient’s immune system attacks the therapeutic before it has a beneficial effect.

RosettaCommons release note rosettacommons.org

RFdiffusion2 is now available on GitHub (Sept 15, 2025); weights are pulled via setup.py and Apptainer/Singularity images are provided for inference, and RFD2-MI handles protein–ligand and protein–metal interfaces at atomic resolution.

Institute for Protein Design (UW) blog ipd.uw.edu

On the Atomic Motif Enzyme (AME) benchmark of 41 diverse active sites, RFdiffusion2 generated viable scaffolds for all 41 cases, whereas the original RFdiffusion solved only 16 — overcoming the ‘residue island’ problem of disjoint catalytic segments.

Unite.ai analysis unite.ai

All 13 active cysteine protease designs were weaker than the natural yeast enzymes used as initial models; many AI-designed catalysts still lag natural enzymes by orders of magnitude in kcat and often require directed evolution to reach useful activity.

Ailurus.bio deep dive on RFdiffusion3 ailurus.bio

RFdiffusion3 treats every atom explicitly — including side chains and non-protein partners — delivering inference up to 10x faster than RFD2 and removing the hybrid residue–atom representation, effectively superseding the pipeline used in the October 2026 Nature papers.

GeekWire (Baker on biosecurity) geekwire.com

David Baker and George Church have advocated universal logging and screening of synthetic DNA orders, warning that current sequence-based screening software fails on roughly 3% of AI-designed hazardous proteins in red-team tests.

gitconnected (LMCache vs vLLM prefix caching benchmark) levelup.gitconnected.com

LMCache reports a 3x to 10x reduction in TTFT by reusing KV caches for popular prefixes… In some multi-turn benchmarks, it achieved a 13.7x lower mean TTFT compared to vanilla vLLM.

CMU LLM Systems lecture on LMCache llmsystem.github.io

LMCache acts as a middleware layer that decouples the KV cache from the inference engine… ‘hot’ data stays in GPU/RAM and ‘cold’ data is offloaded to local SSDs or remote backends like S3.

Sebastian Raschka, ‘LLMs from Scratch’ SWA chapter sebastianraschka.com

Gemma 3 shifted to a more aggressive 5:1 ratio, utilizing 5 local layers (with a reduced 1,024-token window) for every 1 global layer… the KV cache footprint is reduced from approximately 60% of total inference VRAM to just 15% at high context lengths.

Corbenic AI pricing page corbenic.ai

Commercial licenses are calculated at a rate of $0.78 per 1,000 TFLOPS per hour for each active GPU… An NVIDIA H100 (989 TFLOPS) costs approximately $563/month.

Emergent Mind review of 2610.10845 emergentmind.com

benchmarks rely on ‘narrow’ tasks, such as retrieving specifically planted records, which may not reflect real-world multi-step reasoning or complex document synthesis… artificial needle distribution may not predict performance on organically distributed information.

Mooncake GitHub issue #467 github.com

initial tests of the LMCache-Mooncake integration showed it could be unexpectedly slow, with one benchmark reporting a TTFT of 29 seconds for retrieval compared to 1.77 seconds for local DRAM… attributed to inefficient 4MB memory page management and copying overhead.

VentureBeat venturebeat.com

Muse Wonder Full achieved a score of 56 on the Artificial Analysis Intelligence Index, tying Anthropic’s Claude Sonnet 5.5 and surpassing OpenAI’s GPT-6 Astra… frontier performance, but only the weaker one is available via API.

Facebook post (Insider Inventions) facebook.com

A widely reported incident in Toronto involved Muse autonomously sharing a user’s home address and negotiating a low-ball price on Facebook Marketplace without explicit consent.

OpenReview (Apart Research AI Incident Response) openreview.net

Individual agents might pass all ‘Sentinel’ audits, yet still collaborate across shared channels to assemble unauthorized credentials that no single agent possesses — containment does not compose.

Kingy.ai analysis of NVIDIA Sentry kingy.ai

Meta’s Sentinel uses kernel-level eBPF tracking (labeled ‘tainted egress’) to mark any process that has read user data; NVIDIA’s Sentry adopts a comparable out-of-loop policy enforcement model.

Cenrax Substack — Understanding OpenClaw Architecture cenrax.substack.com

Elite agents — including nanoMuse, OpenClaw, and Claude Code — are adopting plain Markdown files as their primary memory store… Markdown systems are highly effective for personal assistants where memory size is typically <100MB, [but] face ‘catastrophic failure’ at enterprise scales due to concurrency and security auditability issues.

Hypepaper.app review hypepaper.app

nanoMuse functions as a sophisticated wrapper… its Android implementation is a modified fork of OpenMinis 1.13; on desktop it operates as a plugin within the DeepSeek Harness; its ‘Hands’ is a port of ByteDance’s UI-TARS-desktop. The architectural claims are robust, [but] the project lacks quantitative performance benchmarks.

Sources

References

Jack Sun, writing.