5 96 104

Shyam Sunder Kumar

theainerd

AI & ML interests

Natural Language Processing

Recent Activity

liked a model 4 days ago

deepseek-ai/DeepSeek-R1-0528

upvoted a collection 4 days ago

MobileLLM

reacted to codelion's post with 🚀 5 days ago

🧠 We just implemented Andrej Karpathy's "third paradigm" for LLM learning! System Prompt Learning (SPL) enables LLMs to automatically learn problem-solving strategies from experience, rather than relying on static prompts. 🚀 How it works: Your LLM builds a database of effective strategies, selects the best ones for each problem, and refines them over time based on success rates. 📊 Results across math benchmarks: Arena Hard: 29% → 37.6% (+8.6%) AIME24: 23.33% → 30% (+6.67%) OptILLMBench: 61% → 65% (+4%) The best part? All strategies are human-readable and the system gets progressively better at problem types you use frequently. ✨ Key benefits: 🔄 Cumulative learning over time 📖 Transparent, inspectable strategies 🔌 Works with any OpenAI-compatible API ⚡ Simple integration: just add "spl-" prefix to your model Built as an open-source plugin in optillm. After 500 queries, our system developed 129 strategies and refined 97 of them! This feels like a genuine step toward AI that learns from experience while staying completely interpretable. 🔗 GitHub: https://github.com/codelion/optillm/tree/main/optillm/plugins/spl 📖 Full article: https://huggingface.co/blog/codelion/system-prompt-learning 🐦 Original Karpathy tweet: https://x.com/karpathy/status/1921368644069765486 Have you experimented with advanced system prompting? What strategies would you want your LLM to learn?

View all activity

Organizations

theainerd's activity

liked a model 4 days ago

deepseek-ai/DeepSeek-R1-0528

Text Generation • Updated 9 days ago • 74.7k • • 1.8k

liked a model 5 days ago

deepseek-ai/DeepSeek-R1-0528-Qwen3-8B

Text Generation • Updated 8 days ago • 139k • • 686

liked 2 models 9 days ago

meta-llama/Llama-Guard-4-12B

Image-Text-to-Text • Updated Apr 29 • 64.8k • 41

fdtn-ai/Foundation-Sec-8B

Text Generation • Updated May 1 • 17.1k • 187

liked a model 16 days ago

mistralai/Devstral-Small-2505

Text2Text Generation • Updated 11 days ago • 183k • 738

liked a dataset about 1 month ago

nvidia/Llama-Nemotron-Post-Training-Dataset

Viewer • Updated 29 days ago • 3.91M • 12.4k • 497

liked 2 models about 1 month ago

deepseek-ai/DeepSeek-Prover-V2-671B

Text Generation • Updated Apr 30 • 7.73k • • 795

Qwen/Qwen3-235B-A22B

Text Generation • Updated 17 days ago • 187k • • 928

liked a Space about 1 month ago

CyberSecEvalTest

📈

Evaluate LLM cybersecurity risks

liked a dataset about 2 months ago

nvidia/ClimbMix

Viewer • Updated Apr 22 • 355M • 5.95k • 30

liked a Space 2 months ago

147

Try YourBench!

🪄

Generate a custom benchmark from any document

liked 4 models 2 months ago

liked a dataset 3 months ago

SakanaAI/AI-CUDA-Engineer-Archive

Viewer • Updated Feb 20 • 30.6k • 760 • 145

liked a Space 3 months ago

375

Reward Bench Leaderboard

📐

Display and filter reward model evaluation data

liked a model 3 months ago

google/gemma-3-27b-it

Image-Text-to-Text • Updated Mar 21 • 404k • • 1.42k

liked a Space 3 months ago

424

AI Deadlines

⚡

Manage project deadlines with AI assistance

liked a dataset 3 months ago

facebook/natural_reasoning

Viewer • Updated Feb 21 • 1.15M • 1.53k • 504