📄 文章列表

1426 篇文章
TDS

Could On-Policy Self-Distillation Revolutionize Post-Training?

✅ completed ⭐ ☆☆☆☆☆ 2026-10-22
VentureBeat

Researchers made AI agents better by turning academic papers into working code — no retraining required

✅ completed ⭐ ☆☆☆☆☆ 2026-10-21
TDS

Skill or MCP Server? Ask What Stops the LLM From Disobeying

✅ completed ⭐ ☆☆☆☆☆ 2026-10-21
TDS

Can Conformal Prediction Take You from Accurate to Reliable Machine Learning?

✅ completed ⭐ ☆☆☆☆☆ 2026-10-20
TDS

Welcome to Google Gemini 4 Argon

✅ completed ⭐ ☆☆☆☆☆ 2026-10-20
TDS

What's in the Box? Distinguishing Acceptable Traces from Unacceptable Ones

✅ completed ⭐ ☆☆☆☆☆ 2026-10-20
TDS

The Hidden Cost of LLM Routing: Why Switching Models Costs More

✅ completed ⭐ ☆☆☆☆☆ 2026-10-20
VentureBeat

Scaling AI agents requires three things: Inventory, contracts and maintained context

✅ completed ⭐ ☆☆☆☆☆ 2026-10-20
VentureBeat

Anyone can build apps with AI. But who owns them?

✅ completed ⭐ ☆☆☆☆☆ 2026-10-19
VentureBeat

DeepSeek-V4.1-Flash-1017 arrives with improved benchmarks just as the U.S. government bans it from agencies — what enter

✅ completed ⭐ ☆☆☆☆☆ 2026-10-19
TDS

AI Harms and Their Technical Causes

✅ completed ⭐ ☆☆☆☆☆ 2026-10-19
TDS

How Many EEG Channels Do You Actually Need?

✅ completed ⭐ ☆☆☆☆☆ 2026-10-19
TDS

Both Charts Are Right. So Why Do They Disagree?

✅ completed ⭐ ☆☆☆☆☆ 2026-10-18
TDS

When EEG Models See the Same People in Training and Testing

✅ completed ⭐ ☆☆☆☆☆ 2026-10-17
VentureBeat

Up to 60% of GPU compute goes to recomputing context. Flash storage can put that capacity back to work

✅ completed ⭐ ☆☆☆☆☆ 2026-10-16
VentureBeat

17 of 40 teams that say they enforce AI agent permissions still share the credentials behind them

✅ completed ⭐ ☆☆☆☆☆ 2026-10-16
VentureBeat

A new framework teaches AI models to audit and repair each other’s reasoning, improving accuracy

✅ completed ⭐ ☆☆☆☆☆ 2026-10-16
TDS

Jev Your Engines: Seven Open Decision Models vs. a Generative LLM

✅ completed ⭐ ☆☆☆☆☆ 2026-10-16
VentureBeat

Meta unveils Muse Spark Open, the new top U.S. open weights model — and it runs on two paired Mac desktops

✅ completed ⭐ ☆☆☆☆☆ 2026-10-15
VentureBeat

NVIDIA's new, open Nemotron 3.5 Super VL with post-training beats Claude Opus 5, GPT-6.1 Sol on specialized task benchma

✅ completed ⭐ ☆☆☆☆☆ 2026-10-15