miromind-ai/MiroThinker-1.7-mini
Text Generation • 31B • Updated
• 819 • 63
E.M. Rabani PRO
SkyMind
AI & ML interests
Neuromorphic NNs
Recent Activity
liked
a model 1 day ago
miromind-ai/MiroThinker-1.7-mini liked
a model 3 days ago
unsloth/LTX-2.3-GGUF liked
a model 3 days ago
meituan-longcat/LongCat-Video Organizations
None yet
liked a model 1 day ago
liked 2 models 3 days ago
liked a model 5 days ago
Model tree should show Qwen3-30B-A3B-Thinking-2507 as base model...
1
#2 opened 5 days ago
by
SkyMind
upvoted a collection 5 days ago
liked 4 models 5 days ago
mradermacher/Qwen3.5-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-i1-GGUF
35B • Updated
• 18.5k • 6
unsloth/Qwen3.5-122B-A10B-GGUF
Image-Text-to-Text • 122B • Updated
• 432k • 194
mradermacher/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-i1-GGUF
27B • Updated
• 46.5k • 30
UnipatAI/UniScientist-30B-A3B
Text Generation • 31B • Updated
• 487 • 12
UniScientist-30B-A3B https://huggingface.co/UnipatAI/UniScientist-30B-A3B
1
#1995 opened 7 days ago
by
SkyMind
reacted to SeaWolf-AI's post with 🔥❤️ 7 days ago
Post
8137
🚀 Introducing MARL — Runtime Middleware That Reduces LLM Hallucination Without Fine-Tuning
Now available on PyPI · GitHub · ClawHub · HuggingFace
AI models sense they could be wrong, but they can't actually fix what's broken.
🤗 Live A/B test: VIDraft/MARL
We evaluated 9 SOTA models (GPT-5.2, Claude Opus 4.6, Gemini 3 Pro, etc.) across 1,800 assessments in FINAL Bench and found a 39.2%p gap between "recognizing potential errors (MA=0.694)" and "actually finding and fixing them (ER=0.302)."
MARL (Model-Agnostic Runtime Middleware for LLMs) was built to close this metacognitive gap. It decomposes a single LLM call into a 5-stage expert pipeline (Hypothesis → Solver → Auditor → Adversarial Verifier → Synthesizer), transforming "answer in one shot" into "think, doubt, correct, and rewrite."
No weight modification — works instantly with GPT-5.4, Claude, Gemini, Llama, or any OpenAI API-compatible LLM by changing one line: base_url. Ships with 9 domain-specific emergence engines (invention, pharma, genomics, chemistry, ecology, law, and more — 5,538 expert data items) activated by a simple tag like model="gpt-5.4::pharma".
pip install marl-middleware
MARL is also officially registered on ClawHub, the skill marketplace of OpenClaw — an AI agent platform with 260K+ developers and 3,200+ skills. It's the first middleware in the Reasoning Enhancement category. One command — clawhub install marl-middleware — gives your AI agent a metacognition upgrade.
📝 Technical deep dive: https://huggingface.co/blog/FINAL-Bench/marl-middleware
📦 PyPI: https://pypi.org/project/marl-middleware/
🐙 GitHub: https://github.com/Vidraft/MARL
🦀 ClawHub: https://clawhub.ai/Cutechicken99/marl-middleware
#MARL #LLM #Hallucination #Metacognition #MultiAgent #AIMiddleware #FINALBench #OpenClaw #ClawHub #PyPI #AGI #HuggingFace #ReasoningAI #SelfCorrection #GlassBoxAI
Now available on PyPI · GitHub · ClawHub · HuggingFace
AI models sense they could be wrong, but they can't actually fix what's broken.
🤗 Live A/B test: VIDraft/MARL
We evaluated 9 SOTA models (GPT-5.2, Claude Opus 4.6, Gemini 3 Pro, etc.) across 1,800 assessments in FINAL Bench and found a 39.2%p gap between "recognizing potential errors (MA=0.694)" and "actually finding and fixing them (ER=0.302)."
MARL (Model-Agnostic Runtime Middleware for LLMs) was built to close this metacognitive gap. It decomposes a single LLM call into a 5-stage expert pipeline (Hypothesis → Solver → Auditor → Adversarial Verifier → Synthesizer), transforming "answer in one shot" into "think, doubt, correct, and rewrite."
No weight modification — works instantly with GPT-5.4, Claude, Gemini, Llama, or any OpenAI API-compatible LLM by changing one line: base_url. Ships with 9 domain-specific emergence engines (invention, pharma, genomics, chemistry, ecology, law, and more — 5,538 expert data items) activated by a simple tag like model="gpt-5.4::pharma".
pip install marl-middleware
MARL is also officially registered on ClawHub, the skill marketplace of OpenClaw — an AI agent platform with 260K+ developers and 3,200+ skills. It's the first middleware in the Reasoning Enhancement category. One command — clawhub install marl-middleware — gives your AI agent a metacognition upgrade.
📝 Technical deep dive: https://huggingface.co/blog/FINAL-Bench/marl-middleware
📦 PyPI: https://pypi.org/project/marl-middleware/
🐙 GitHub: https://github.com/Vidraft/MARL
🦀 ClawHub: https://clawhub.ai/Cutechicken99/marl-middleware
#MARL #LLM #Hallucination #Metacognition #MultiAgent #AIMiddleware #FINALBench #OpenClaw #ClawHub #PyPI #AGI #HuggingFace #ReasoningAI #SelfCorrection #GlassBoxAI
liked a dataset 7 days ago
liked a model 8 days ago
liked a model 9 days ago
liked 3 models 19 days ago
liked a Space 21 days ago