A Coding Guide to Google Research’s MSEB: Writing Sound Encoders to the Benchmark Contract and Scoring Them Across Classification, Clustering, Retrieval and Segmentation
15 min
Claude Opus 5.5 uses 95% fewer em dashes, but its answers are getting longer
2 min
OpenAI's GPT-6 Astra can now tell you exactly where you screwed up your IKEA shelf
Coding Agents for Generalized Task and Motion Planning Problems
To Trust or Not to Trust: Retrieval-Augmented Fact Checking in Speech
A Living Benchmark for Information Retrieval from Electronic Health Records
ExplorationBench: Measuring AI Systems' Exploration in Verifiable Alien Worlds
PrivDrift: Auditing User-Secret Leakage Under Topic Drift in Active LLM Conversations
AI performance costs are falling faster than those of any previous technology
2 sources 4 min
Just Ask Jev: Reinforcement Learning for Calibrated Decisions as a Zero-Shot Detector of AI Alignment Failures
Mercury 2.5 LLM hits 770 tokens per second
6 min
RGBD20K: A Large-Scale Benchmark for RGB-D Semantic Segmentation
StudentBench: AI and human tutoring yield equivalent GRE learning gains
Can LLMs Reason About Runtime Behavior? A Repository-Level Dynamic Benchmark
GPT-6 Astra has gained the ability to drive a car
2 min
545 Hackers Tested It First. Now XRanges for AI Scores Your Security Agent
8 min
Introducing MentalHealthBench
8 min
LLM Ass Bench
EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics
WhatWorkedBench: Benchmarking Experimental Understanding in AI Agents
SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving
Evaluating Coding Agents on Kernel Exploit Generation
1
RoboFollow: Unveiling the Instruction Following Mirage in Embodied Agents
