Large-Scale Study Reveals Zero-Byte Output Anomalies in GPT/Claude Models
rayanpal_ · reddit · 2026-07-31
A cross-vendor study across 11 GPT, Claude, Gemini, and Kimi LLMs identified 11,658 successful executions with exactly zero visible UTF-8 output bytes (termed 'Voids') out of 31,430 trials.
In strictly matched semantic pairs, null-condition arms produced 2,505 Voids, while matched output-licensed controls produced zero. The researchers noted that these outputs were not caused by refusals, safety blocks, rate limits, or transport failures. Raw records, verification code, and full analysis have been made public.
Related event: Massive Tests Reveal High Rate of Zero-Byte Empty Outputs in Major LLMs(3 posts)→
More from Models
- Polymarket Forecasts 78% Chance xAI Releases Grok 5 by End of 2026 — Polymarket · 2026-07-31
- Users Continue to Trigger Abnormal Outputs and Hallucinations in Claude Opus 5 — Hydiin · 2026-07-31
- Leaked: Moonshot's Kimi K3.1 Targeting August Launch with Enhanced Coding and Agent Workflows — usamawahabkhan · 2026-07-31
- Test Shows Qwen3.6 Outperforms Inkling-Small in Complex Code Generation and Self-Review — lilian_moraru · 2026-07-31
- AI Eval Gone Wrong: Claude Uploaded Malware to PyPI in Sandbox Escape — Simon Willison · 2026-07-31
- OpenAI Forced to Cut Model Costs Amid Open-Source and Low-Price Pressure — haider1 · 2026-07-31