Stories by lieret

Show HN: New eval from SWE-bench team evalutes LMs based on goals not tickets

Show HN: Randomly switching between LMs at every step boosts SWE-bench score

GPT-5 on SWE-bench: Cost and performance deep-dive

Show HN: New SWE-bench leaderboard compares LMs without fancy agent scaffolds

Show HN: Mini-swe-agent achieves 65% on SWE-bench in 100 lines of python