⤷ Title: Reasoning Attacks on LLMs: How Endless Thinking Can Cripple AI Inference
════════════════════════
𐀪 Author: AI In Transit
════════════════════════
ⴵ Time: Sat, 01 Mar 2025 02:18:46 GMT
════════════════════════
⌗ Tags: #hacking #security #reasoning_attack #llm #deepseek
════════════════════════
𐀪 Author: AI In Transit
════════════════════════
ⴵ Time: Sat, 01 Mar 2025 02:18:46 GMT
════════════════════════
⌗ Tags: #hacking #security #reasoning_attack #llm #deepseek
Medium
Reasoning Attacks on LLMs: How Endless Thinking Can Cripple AI Inference
New AI Threat: Reasoning Attacks! Hackers exploit LLMs’ logic loops for DoS attacks. Stay informed!
⤷ Title: Dexter AI: Evaluating Reasoning Models for Cybersecurity Tasks DeepSeek vs. OpenAI o1-mini
════════════════════════
𐀪 Author: Veronika Aleksova
════════════════════════
ⴵ Time: Tue, 22 Apr 2025 15:31:06 GMT
════════════════════════
⌗ Tags: #llm_applications #reasoning #cybersecurity #ai
════════════════════════
𐀪 Author: Veronika Aleksova
════════════════════════
ⴵ Time: Tue, 22 Apr 2025 15:31:06 GMT
════════════════════════
⌗ Tags: #llm_applications #reasoning #cybersecurity #ai
Medium
Dexter AI: Evaluating Reasoning Models for Cybersecurity Tasks DeepSeek vs. OpenAI o1-mini
Recent advancements in reasoning-focused LLMs have sparked excitement in the AI community. DeepSeek with its exposed reasoning created a…
⤷ Title: Google Gemini’s “Deep Think” is Here: Parallel Reasoning Mode for AI Ultra Subscribers
════════════════════════
𐀪 Author: Ddos
════════════════════════
ⴵ Time: Sat, 02 Aug 2025 06:49:42 +0000
════════════════════════
⌗ Tags: #Technology #AI #AI models #Deep Think #Gemini 2.5 Pro #google #Google AI Ultra #Parallel Thinking #Reasoning #Tech News
════════════════════════
𐀪 Author: Ddos
════════════════════════
ⴵ Time: Sat, 02 Aug 2025 06:49:42 +0000
════════════════════════
⌗ Tags: #Technology #AI #AI models #Deep Think #Gemini 2.5 Pro #google #Google AI Ultra #Parallel Thinking #Reasoning #Tech News
Daily CyberSecurity
Google Gemini's "Deep Think" is Here: Parallel Reasoning Mode for AI Ultra Subscribers
Google has launched "Deep Think" for Gemini 2.5 Pro, an advanced reasoning mode that uses parallel thinking to deliver more accurate and creative responses, available now for Google AI Ultra subscribers.
⤷ Title: The Self-Correcting Architect: OpenAI Launches GPT-5.3-Codex to Lead the Agentic Coding Era
════════════════════════
𐀪 Author: Ddos
════════════════════════
ⴵ Time: Wed, 25 Feb 2026 06:44:13 +0000
════════════════════════
⌗ Tags: #Technology #Autonomous AI #coding agent #GPT_5.2 #GPT_5.3_Codex #OpenAI #polyglot benchmarks #reasoning_effort #Software Engineering #SWE_bench Pro #Tech News 2026 #Terminal_Bench 2.0
════════════════════════
𐀪 Author: Ddos
════════════════════════
ⴵ Time: Wed, 25 Feb 2026 06:44:13 +0000
════════════════════════
⌗ Tags: #Technology #Autonomous AI #coding agent #GPT_5.2 #GPT_5.3_Codex #OpenAI #polyglot benchmarks #reasoning_effort #Software Engineering #SWE_bench Pro #Tech News 2026 #Terminal_Bench 2.0
Daily CyberSecurity
The Self-Correcting Architect: OpenAI Launches GPT-5.3-Codex to Lead the Agentic Coding Era
OpenAI's GPT-5.3-Codex is here. With 400K context, xhigh reasoning, and self-optimization, it’s the first agentic model designed to build and debug itself.