PROOF-Gen: From Optimized Data to Better Distillation
Supervised fine-tuning on teacher-generated trajectories is the standard first stage for distilling tool-calling capabilities into deployable models. Post-training pipelines that drive shipped tool-calling agents re-run this stage on a daily or weekly cadence, paying the frontier-teacher cost each c...
IDEA Prune: An Integrated Enlarge-and-Prune Pipeline in Generative Language Model Pretraining
Recent advancements in large language models have intensified the need for efficient and deployable models within limited inference budgets. Structured pruning pipelines have shown promise in token efficiency compared to training target-size models from scratch. In this paper, we advocate incorporat...
Luce: Relightable Gaussians for 3D Asset Generation
High-fidelity image-to-3D generation requires a 3D representation that captures both geometry and appearance. To support relighting and integration into standard rendering pipelines, the representation should include physically based rendering (PBR) modalities such as albedo, metallic-roughness, and...
Liquid AI Open-Sources Pipette: A Reproducible Benchmarking Suite That Measures On-Device Models, Quantization, Runtime and Hardware Together
Model cards report quality under server-class, full-precision conditions. Those numbers rarely predict how the same model behaves on a phone. This week, Liquid AI released Pipette. It is an open-source platform for benchmarking foundation models on edge devices, built in partnership with Artificial ...
Perplexity Ships Portable Computer on NVIDIA DGX Spark: Local Harness, OS-Enforced Sandbox, and Zero Per-Token Cost for Local Steps
Perplexity releases Portable Computer, packaging local models, harness, sandbox, and connectors into one system running on NVIDIA DGX Spark.
The post Perplexity Ships Portable Computer on NVIDIA DGX Spark: Local Harness, OS-Enforced Sandbox, and Zero Per-Token Cost for Local Steps appeared first on ...
Meta AI Introduces MetaRoCE: A Clean-Sheet RDMA Transport Built for AI-Scale Ethernet
Training and serving frontier models is now a networking problem as much as a compute problem. Collective operations like all-reduce and all-to-all synchronize thousands of accelerators during training, and the slowest transfer sets the pace for the entire job. Even small amounts of network friction...
A New Towards Data Science: A Faster Site and a Brand-New Contributor Portal
We're excited to share some big news: We completely rebuilt the TDS website and our contributor portal. Whether you come here to read, to write, or both, here's what you can expect from our new site.
The post A New Towards Data Science: A Faster Site and a Brand-New Contributor Portal appeared first...
The Design System as the Control Plane for AI-Generated UI
AI-assisted development has made it easier to generate frontend code quickly. A developer can ask for a form, a dashboard widget, a settings page, or a modal flow and get a working first draft in seconds. That speed is useful, especially when teams are moving through routine UI work. But speed creat...
OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show
Tested on Semianalysis’s InferenceX benchmark, Jalapeño registered both more tokens per user and more throughput per kilowatt than the currently available state-of-the art.
Hallucinations, Watermarks, Removers, and a Squeezed Balloon
Watermarks act at the model’s moments of doubt, and so do the safety checks that catch AI mistakes
The post Hallucinations, Watermarks, Removers, and a Squeezed Balloon appeared first on Towards Data Science.
Shadow Agents, Standing Privileges, and the Governance Gap Between Deployment and Discovery
There was a brief window where AI agent security felt like a future problem. Organizations deployed copilots, coding assistants, and autonomous workflows on the assumption that the worst case was a bad recommendation or a hallucinated answer. That window closed in the first half of 2026, when a clus...
A practical guide to navigate hierarchies, find routes, detect cycles and calculate degrees of separation
The post Recursive CTEs: SQL’s Hidden Graph Traversal Engine appeared first on Towards Data Science.
I spent a day at a robot “carnival” in Shanghai. Here’s what I saw.
Humanoid robots are having a moment in China. The popular machines are part of the country’s strategy to bring artificial intelligence into daily life. Embedding the technology into physical systems—an idea called embodied AI—was a key facet of China’s latest five-year plan, and companies here are a...
OpenAI CFO Sarah Friar explains how advances across chips, compute, models, and products compound to deliver more useful intelligence at greater scale and lower cost.
Jalapeño’s first results show industry-leading speed and efficiency in AI inference
Jalapeño is a custom inference chip from OpenAI that delivers faster, more power-efficient AI inference, with higher throughput and lower latency for modern models.
AI Learning and Conceptual Transfer in the Game of Hidden Rules
arXiv:2608.21372v1 Announce Type: new
Abstract: This report summarizes the work conducted on the Game of Hidden Rules (GOHR), focusing on reinforcement learning agents trained to infer hidden rules from trial-and-error feedback, representation design, rule difficulty analysis, transfer learning, ge...
STARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal Generation
Unified multimodal models that understand, reason over, and generate interleaved text–image sequences remain structurally fragmented: existing approaches either sacrifice visual fidelity through discrete tokenization, impose structural asymmetry by combining causal text generation with iterative dif...