techxiv
Defining and evaluating political bias in LLMs 
Server rendering benchmarks: Fluid Compute and Cloudflare Workers 
From Single-Node to Multi-GPU Clusters: How Discord Made Distributed Compute Easy for ML Engineers 
Kurt Got Got 
Deploy Safety: Reducing customer impact from change - Disrupting malicious uses of AI: October 2025
Improving Java support in Cursor 
Accelerating our Android apps with Baseline Profiles 
Inside Husky’s query engine: Real-time access to 100 trillion events - Cyber Operation: Russian-speaking malware tooling
- Cyber Operation: Phishing and scripting support
- Operation “Nine–emdash Line”: Regional influence activity
- Operation “Stop News”: Recidivist influence activity
- PRC-linked abuse: Surveillance and influence activity
- Cyber Operation: Korean-language malware support
- Scam operations: Online fraud networks
- LLMs Are the Key to Mutation Testing and Better Compliance
- Sora 2 System Card
- Turning contracts into searchable data at OpenAI
- Meta’s Infrastructure Evolution and the Advent of AI