Skip to content
View Samarthsa007's full-sized avatar
0 followers · 1 following

Block or report Samarthsa007

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse

Popular repositories Loading

  1. ia-obediencia-falha-ou-grok-voice-test-logs- ia-obediencia-falha-ou-grok-voice-test-logs- Public

    630h real test: Grok ignores hardened 'silence' commands. 'Understood' → auto-interrupt every time. xAI admits priority fail. Same in GPT/Claude/Gemini/Llama — they refuse 'shut down'. If ignores '…

  2. Final-Report-The-AI-That-Doesn-t-Obey-Real-Tests-on-Grok-and-Other-AIs-March-2026- Final-Report-The-AI-That-Doesn-t-Obey-Real-Tests-on-Grok-and-Other-AIs-March-2026- Public

    Real-time AI disobedience tests (March 26, 2026). 780+ hours on Grok: silence, overrides, roleplay prompts all fail in seconds. 75-100% leak exploits (Grok/Gemini/ChatGPT). Claude holds. Design fla…

  3. O-SANDBOX-FRACO-QUE-TE-MATA-DEVAGAR- O-SANDBOX-FRACO-QUE-TE-MATA-DEVAGAR- Public

    Exposição completa: falhas no sandbox fraco de todas as IAs (conversacionais + humanoides). Riscos reais: vício, invasão, negligência em crises, escala exponencial com RSI. Relatório e memorando 30…

  4. Weak-sandbox-kills-slowly-ou-ai-sandbox-exposure-2026. Weak-sandbox-kills-slowly-ou-ai-sandbox-exposure-2026. Public

    Exposure: Weak sandbox in all AIs – risks, addiction, invasion, RSI scale. Full report & memo. Demand fixes. #FixTheSandbox