Sulbha Jain·Nov 3Breaking the Sanskrit Code: How Ancient Grammar Meets Modern Logic!Have you ever tried reading an old Sanskrit text and felt like all the words were crammed together into one giant, confusing puzzle? That’s…
Sulbha Jain·Mar 8You’re Probably Using Cosine Similarity Wrong ; And It’s Quietly Breaking Your RAG PipelineA look at a 2024 Netflix/ACM paper that shook my assumptions, and how I rebuilt retrieval from scratch using a multi-embedding approach.
Sulbha Jain·Mar 8Building a Hotel Booking AI Agent with GRPO and Live Database ToolsHow I used Reinforcement Fine-Tuning to teach Qwen3–0.6B to call real PostgreSQL functions — accurately
Sulbha Jain·Mar 6Why Agentic RL Breaks (and How rStar2-Agent Fixes It) — Paper ReviewIf you’ve ever watched an LLM use tools during reasoning, you’ve seen the magic:
Sulbha Jain·Mar 5Your RAG System Is Paying a Tax It Doesn’t Owe: REFRAG — Paper ReviewPicture this: your AI assistant gets a question. It retrieves ten relevant documents, stuffs them into the context window, and sends the…
Sulbha Jain·Mar 3Why I Stopped Using GPT-4 for 70% of My AI Agent Calls? — Paper ReviewAnd Why You Should Too
Sulbha Jain·Feb 25Stop Sending Every Query to GPT-4A Practical Guide to Budget-Aware LLM Routing with PILOT — Paper Review
Sulbha Jain·Feb 23The Quiet Revolution in AI ReasoningWhy the Best Thinking Happens in Silence — Paper Review
Sulbha Jain·Feb 23Why Your AI Confidently Lies to You?And Why That’s Actually a Math Problem — Paper ReviewA response icon1A response icon1