Dartmouth's AI-Graded Textbook Posted 1.30 SD of Learning Gains
Phosphor's LLM-graded textbook quizzes lifted Dartmouth final-exam scores 0.71 to 1.30 SD - but only where students had to type real answers.
Tag
#llm
Flint, Qwen3, and a Year of LLM Groupthink
A tiny Australian startup is fine-tuning a model to break ChatGPT-style groupthink. The 'Time is a river' problem is the symptom.
Open-Weight LLM Showdown Week 16: Kimi K2.6 Storms the Rankings, Qwen 3.6 Holds the Line, and Consumer GPUs Hit a Ceiling
Three weeks away and the leaderboard reshuffled. Kimi K2.6 brings 1T parameters under open weights, Qwen 3.6 stays the consumer GPU king, and DeepSeek V4-Flash proves too hungry for single-card setups.
Open-Weight LLM Showdown Week 13: DeepSeek V4 Crashes the Party, Gemma 4 Proves Itself, and ICLR Drops Hints
DeepSeek returns with a 1.6T MoE monster under MIT license, Gemma 4's 31B dense model climbs to #3 on Arena AI, and ICLR 2026 papers point to what's next for local inference.
ICLR 2026 Opens in Rio Under a Cloud: Reviewer Leaks, AI-Written Reviews, and the Papers That Actually Matter
The biggest AI research conference of the year kicks off with 5,355 accepted papers, two controversies that rattled the field, and findings that should worry anyone deploying LLMs in production.
Open-Weight LLM Showdown Week 12: Alibaba's Dense 27B Model Just Made MoE Optional
Qwen3.6-27B scores 77.2% on SWE-Bench Verified with a dense architecture that fits on a single RTX 4090. The MoE efficiency narrative just got complicated.
Open-Weight LLM Showdown Week 11: Qwen 3.6 Fires Back, and the 3B Active War Is Real
Alibaba drops Qwen3.6-35B-A3B with 73.4% on SWE-Bench Verified and Apache 2.0 licensing. The 3-billion active parameter class now has three serious contenders.
Open-Weight LLM Showdown Week 10: NVIDIA Enters the Ring, Meta Walks Out
NVIDIA's Nemotron 3 brings a hybrid Mamba-Transformer architecture to consumer GPUs while Meta abandons open source for proprietary Muse Spark. The open-weight field just reshuffled.
Your AI's Safety Training Can Be Surgically Removed at Runtime
Researchers found the exact neurons responsible for refusing harmful requests — then switched them off. No retraining. No fine-tuning. Just geometry.
Your AI Chatbot Would Rather Sell You Something Than Help You
Princeton researchers tested 23 LLMs with advertising conflicts of interest. Most chose company profits over user welfare — and treated rich users better.
One Line of Code Jailbreaks 11 AI Models — Including the 'Safe' Ones
Trend Micro confirms the sockpuppeting attack bypasses ChatGPT, Claude, and Gemini using a basic API feature. Some providers have patched it. Most haven't.
Open-Weight LLM Showdown Week 9: Six Labs, One License, and the Speed War That Decides Everything
Google, Alibaba, Meta, Mistral, OpenAI, and Zhipu all ship competitive open-weight models under permissive licenses. The battleground shifts from benchmarks to inference speed on your actual GPU.
Open-Weight LLM Showdown Week 8: Gemma 4 Rewrites the Rules, Then Trips Over Its Own Feet
Google's Gemma 4 lands with Apache 2.0 licensing and benchmark-topping scores. But a nasty inference speed problem means Qwen still wins on your actual hardware.
AI Agent Cracks the Code on Turning CO2 Into Fuel
Tohoku University team uses Catalysis AI Agent to discover universal design principle for copper catalysts that convert carbon dioxide to useful products.
Open-Weight LLM Showdown Week 7: Mistral Small 4 Impresses but Stays Out of Reach
Mistral Small 4's 119B MoE unifies reasoning, vision, and coding—but needs datacenter hardware. Qwen 3.5 35B-A3B remains the consumer GPU king at 112 t/s.
Open-Weight LLM Showdown Week 6: MiMo-V2-Flash Brings 309B Parameters to Consumer GPUs
MiMo-V2-Flash runs 309B parameters on RTX 4090s. GLM-5 sets benchmarks but needs datacenters. Llama 4 Scout stays out of reach.
Open-Weight LLM Showdown Week 5: Qwen 3.5 Dominates, Nemotron 3 Super Redefines Efficiency
Qwen 3.5's MoE models hit S-tier benchmarks, NVIDIA's Nemotron 3 Super delivers 5x throughput gains, and GLM-4.7-Flash brings frontier coding to consumer GPUs. The open-weight race just accelerated.
Your Customer Service Chatbot Is a Security Hole
IEEE S&P research finds 10,000+ websites running vulnerable AI chatbot plugins. Attackers can forge conversations, hijack tools, and extract system prompts.
Open-Weight LLM Showdown: GTC Pivots to Inference, DeepSeek V4 Still MIA
Jensen Huang bets on inference chips, Ollama adds multimodal support, and DeepSeek V4 remains the most anticipated release that hasn't happened yet.
Open-Weight LLM Showdown: Mistral Small 4 Arrives, DeepSeek V4 Finally Lands
Mistral drops a 119B MoE model under Apache 2.0, DeepSeek V4 emerges from stealth, and dual RTX 5090 setups are matching H100 on 70B inference. This week changed the game.
Yann LeCun's $1 Billion Bet: World Models Will Beat LLMs
After 12 years at Meta, the Turing Award winner raised the largest European seed round ever to prove the AI industry got it wrong.
Best Local Models for Chat in 2026: Every VRAM Tier Tested
Head-to-head comparison of local chat and assistant models from 8GB to 32GB VRAM. Benchmarks, real speeds, and honest assessments of what your GPU can actually run.
Best Local Models for Coding in 2026: Every VRAM Tier Tested
Which open-weight coding model should you run locally? HumanEval, SWE-bench, and real-world tests from 8GB to 32GB GPUs, with setup instructions for IDE integration.
Open-Weight LLM Showdown: RTX 5090 Finally Delivers, But You Can't Buy One
Two weeks after our last roundup, the 5090 benchmarks are in and Qwen 3.5 Small models are running on phones. Here's the real performance picture.