The Archive

Traversal, an AI agent platform for SRE automation, raised $120M at $1B valuation from KP, Redpoint, Khosla.

llm 0.37 adds GPT-6.1 variants, OpenAI-compatible endpoint config, defaults to cheaper GPT-6-Luna.

Amazon Bedrock AgentCore Session endpoints give each runtime session an authenticated URL that forwards any HTTP method, path, and WebSocket connection to the ports you allow. In this post, you deploy a coding agent that keeps a Vite website running in its session, then watch the live site update in your browser as the agent edits files.

Teams building AI applications need a repeatable path from a model configuration to a working inference endpoint. NVIDIA Dynamo is a distributed inference... Teams building AI applications need a repeatable path from a model configuration to a working inference endpoint. NVIDIA Dynamo is a distributed inference framework that coordinates model serving across GPU workers. Deploying it on Kubernetes is only a first step. A successful model deployment requires more: GPU drivers, resource allocation, scheduling, networking, and monitoring must all work… Source

Reliability-Weighted DPO improves LLM alignment robustness by learning weights for noisy preference comparisons.

MoEngage turned roughly 400 private REST APIs across 7 AWS Regions into permission-scoped agent tools with Amazon Bedrock AgentCore Gateway. No per-service MCP servers, no public endpoints, and no 400-tool context bloat. See how OpenAPI specs, Cedar policies, and managed Amazon VPC Lattice egress made it work.

Studies solution-set structure of grokking on Latin square tasks, characterizing which minimum-norm endpoints enable generalization from partial observations.

Trajectory Matching: discrete flow matching supporting structured multi-component actions (fragments, nodes, hierarchies) via endpoint conditioning.

Curate-DTI curates 2.9M drug-protein bioactivity measurements with construct-aware protein targets for improved ML dataset quality.

Systematic study of DPO hyperparameter landscape across model scales reveals optimal learning rate and β-regularization shift with size; direct transfer between scales unreliable.

Waypoint-OPD improves policy distillation by tracking intermediate policy shifts during RL training, not just endpoint differences.

now supports MCP Events in public beta, allowing your apps and agents to subscribe to webhook events from compatible MCP servers. Build agents for tasks like summarizing new messages, reviewing newly created issues, or responding to document feedback.Vercel Connect Connect creates subscriptions on the MCP server and renews them automatically as needed, verifying the server’s webhooks and forwarding events to your application using authenticated delivery. You can use the Connect SDK to subscribe to MCP events: At your delivery endpoint, use the SDK to verify that each delivery came from Connec...

Manacá-1B-Instruct-2 Brazilian Portuguese LLM with documented negative DPO results and analysis of LLM judge failure modes in safety alignment.

Graph-enhanced framework discovers 99,527 disease trajectory motifs from MIMIC-IV with leakage-controlled splits and time-aware Transformer endpoint forecasting.

JIVE training-free framework enhances generative sample diversity via Jacobian-informed velocity perturbations, provably maximizing endpoint diversity without mode collapse.

End-to-end driving models suffer open-loop/closed-loop training mismatch; endpoint-constrained trajectory optimization improves physical coherence.

The zero-day exploit required local access to the user’s device, but gave potential attackers access to Muse accounts. | Image: The Verge Meta has issued a patch for its Muse macOS app following the discovery of a zero-day vulnerability that could allow someone to take control of the AI agent. The bug found by security researcher Patrick Wardle utilized an undocumented Muse setting that enabled potential attackers running local code to redirect transcription processing from Meta's servers to their own endpoint, Ars Technica reports, giving the attacker access to the Muse account. Several desi...

Intrinsic Robot Rewarding reuses VLA frozen encoders and demonstration endpoints to autonomously evaluate and improve robot policy outcomes.

One-step distillation for point cloud cranial implant generation; medical imaging application with limited AI frontier relevance.

OpenRouter's cost-optimization routing across providers causes behavioral inconsistency—same model endpoint yields different outputs due to varying serving software and feature gaps.

Answer-distribution trajectories track full predictive distributions over LLM reasoning steps, revealing competing hypotheses beyond endpoint accuracy.

Preregistered audit of LLM-as-judge reliability finds Spearman 0.40 repeat agreement (vs. 0.90 required), exposing instability in model endpoints underpinning leaderboards and training data filtering.

PLC-DPO corrects mislabeled preference pairs in DPO training by classifying each pair as clean, flip, or tie using policy-reference margin calibration.

ToxLens applies graph learning to molecular toxicity prediction with leakage-aware splitting and calibrated uncertainty across 11 toxicity endpoints.

Tool: CORS Chat I built this today ( with GPT-5.6-Sol xhigh ) to help test Qwen 3.8 27B running in LM Studio on both my M5 MacBook Pro and an NVIDIA DGX Spark. It provides a web UI for exercising an OpenAI-Responses-compatible chat endpoint. I've tried it against LM Studio with the --cors option and OpenRouter, and both work fine. Conversations are persisted in the browser and can be exported as copy-pasted JSON. One fun detail is that it notices SVG images that are being generated and progressively renders them in the chat while the tokens are still streaming in. Tags: svg , ai , generative-...

Many clinical prediction models treat post-intervention outcomes as a one-step mapping from baseline measurements to a future endpoint. However, recovery after a procedure often unfolds as an irregular trajectory: clinical observations, medication changes, repeat interventions, and physiological measurements are recorded asynchronously and can change risk assessment over time. We propose an intervention-aware clinical world model that represents each patient with a structured latent state and evolves it through time-ordered post-intervention events. The model first encodes baseline imaging in...

Foundation models for protein structure prediction remain unreliable on certain targets. External oracles can flag and correct these failures, but biological oracles are expensive, making oracle budget a critical constraint. Existing guidance methods, such as FK-steering, DPO, and Best K-of-N sampling, differ in how they spend this budget, yet no systematic comparison exists to guide method selection. To bridge this gap, we benchmark these methods alongside the recently proposed Optimisation Over Outputs (O3), which applies off-the-shelf optimisers within a generative model's latent subspace....

ReRound (Reconstructive Rounding) is a post-training quantization method that addresses the midpoint ambiguity inherent in standard round-to-nearest (RTN) schemes when quantizing weights near the centers of quantization intervals. Starting from a pretrained LLM, ReRound trains a conditional diffusion model to produce continuous reconstructions of low-bit weights for the LLM. These reconstructed weights act as a guidance signal to disambiguate the rounding direction of weights located close to interval midpoints. To integrate this reconstruction-guided rounding with conventional RTN, ReRound i...