Show HN: Pencil Puzzle Bench – LLM Benchmark for Multi-Step Verifiable Reasoningppbench.com·5 pts·bluecoconut·0
The Dataflow Model: Practical Approach to Balancing Correctness Latency and Costresearch.google·2 pts·bluecoconut·0
Bird-SQL: A Big Bench for Large-Scale Database Grounded Text-to-SQLsbird-bench.github.io·2 pts·bluecoconut·0
Show HN: Sketch – AI code-writing assistant that understands data contentgithub.com·252 pts·bluecoconut·49