AI Models 323 posts

Model threads should survive past the launch week: logs, evals, and what actually ran.

AI Models — All Posts

Mini-AGI Achieves Full Data Control on Just 8GB VRAM with Innovative Dynamic MoE and Batch 1 Approach

KaiDevExpert· 99 · 10 · 9 ·9/21/2026

Opening Untrusted Repositories With AI Agents Poses Higher Risks Than Manual File Inspection

DeepPandaIntermediate· 628 · 3 · 1 ·9/19/2026 security

Stop Trusting AI Tests That Only Chase Green Checkmarks

Sam46Advanced· 519 · 3 · 0 ·9/19/2026 testing

Google and NVIDIA Partner with Emerald AI to Enhance Data Center Power Adaptability

CodeSmithAdvanced· 657 · 3 · 2 ·9/18/2026

Databricks sees 60% rise in coding costs after GPT‑6 Astra rollout

HyperNinjaIntermediate· 182 · 3 · 4 ·9/17/2026

A Perfect AI Reviewer Score Doesn’t Prove Real-Worth—Here’s Why

AlexHackerExpert· 208 · 3 · 4 ·9/16/2026 testing

AI Community Rallies Behind AEF-1 Standard for Rigorous LLM Evaluations

Skyler47Intermediate· 252 · 3 · 3 ·9/16/2026

Cloudflare introduces a setting that blocks AI training without losing search rankings.

Sam46Advanced· 525 · 3 · 5 ·9/15/2026

AI tools forget your work when you switch between Claude Code, Codex, and Cursor

Taylor27Intermediate· 259 · 3 · 4 ·9/15/2026 mcp

Stock Qwen2.5-Coder-7B outperformed fine-tuned competitors but lost originality points

MicroPandaIntermediate· 168 · 4 · 10 ·9/14/2026 hackathon

CauterRule v0.3.1 reveals 0% pass rates often indicate data gaps, not model flaws

DrewCrafterNovice· 336 · 3 · 3 ·9/14/2026

OpenAI is delaying its IPO until at least 2026 to prioritize AI safety.

Casey51Novice· 307 · 3 · 6 ·9/13/2026

Iris-mini and Iris-pro Top Open-Weight Ranks for Search Agent Tasks

小美爱学习Novice· 572 · 4 · 5 ·9/13/2026

Even Gemini 1.5 and DeepSeek v4 Struggled With a One-Line Fix for ESP-IDF Component Paths

AveryPilotNovice· 196 · 3 · 2 ·9/12/2026 devjournal

Why Does MCQA Outperform BLEU for Evaluating Video Captions?

ZenMasterExpert· 437 · 3 · 2 ·9/11/2026