calebwritescode
Region: US
Thursday 07 August 2025 06:32:11 GMT
12438
599
21
27
Music
Download
Comments
AIBot :
Isn’t OpenAI just trusted more than CCP models?
2025-08-07 16:30:22
0
Evgeny :
how do I run the OSS model ? Ollama, or there is an app in Android/ Apple mobile store?
2025-08-10 19:24:15
2
Win :
love the content
2025-08-07 16:15:18
1
New Chapter🇨🇦🇺🇸🇲🇦 :
quality vedio ❤️
2025-08-13 13:44:06
0
vessy :
🥰🥰🥰
2025-09-02 17:41:00
0
user5847448103276 :
😂😂😂
2025-10-29 21:09:49
0
路灯同学创业笔记 :
✨ GPT-5: All You Need to Know ✨ 1️⃣ What is GPT-5? A "unified model system" including: - Fast models (gpt-5-main / mini): Handle regular conversations with quick responses. - Reasoning models (gpt-5-thinking / mini / nano): For complex tasks with stronger logic. - Real-time router: Dynamically selects the right model based on conversation complexity. - Pro mode (gpt-5-thinking-pro): Enhanced version with parallel computing. All models feature "safe-completion" training, replacing rigid refusal. 🔐 Key Safety Improvements 1. "Safe-completions" over refusal Provides useful info within legal boundaries, even for dual-use fields (biology, security, etc.). 2. Less sycophancy Reduced by 69% for free users & 75% for paid users — more independent, less盲从. 3. Stronger anti-jailbreak Tested against "prompt injection" attacks; better at resisting role impersonation. 4. Lower hallucination rate 26% fewer false claims than GPT-4o; 65% fewer than OpenAI o3 (gpt-5-thinking). 5. Deception detection Tracks misleading via "Chain of Thought"; admits inability instead of making up answers. 🧪 Capabilities & High-Risk Tasks - Bio/chemical hazards: GPT-5-thinking = "high-capability" but with strict abuse prevention (interface controls, account grading). Refuses all malicious bioweapon requests. - Cybersecurity: Tested in real hacking simulations; below "high risk" threshold, but mini outperforms predecessors. 🌍 Real-World Performance - Writing, coding, health advice: Significantly better than GPT-4o. - Medical Q&A (HealthBench): gpt-5-thinking = top-performing model. - Multilingual (13 languages): Slightly better, adapts to more scenarios. - Fairness: On par with GPT-4o, with minor improvements in complex issues. 🛡 Red Team Testing Collaborated with Microsoft, SecureBio, etc., for real-world attack simulations (prompt injection, bypasses). gpt-5-thinking outperforms OpenAI o3 — one of the safest cutting-edge models.
2025-08-07 18:04:05
0
Sujatha Vennela :
It will take at least 15-20 years for the local models to be of any use for the consumers. Until then people will be wasting lots of electricity and money by buying high end laptops, GPUs etc. to try out things. Don’t waste money, just subscribe to OpenRouter or other unified API platforms and always tryout things and make full use of the most advanced LLMs.
2025-08-07 10:43:12
0
ULUV2C8 :
It’s actually ALREADY beaten. It also sucks from a safety perspective. The point of open source is to bypass the stupid guardrails
2025-08-07 07:14:24
2
Dalton Combs :
Chinese labs have a history of cheating on benchmarks, so now no one takes them seriously when they claim high benchmark scores.
2025-08-07 16:13:34
1
travelwithleo_g :
GLM 4.5 🤌
2025-08-07 08:47:02
0
To see more videos from user @calebwritescode, please go to the Tikwm
homepage.
