Zachary Lipton

A tiny LLM just beat the giants at their own game. In a controlled evaluation test this summer, gpt-5-nano consistently picked the right answer while Claude-Sonnet-4 and Gemini 2.5 Pro fell for a reasoning trap. The smaller model not only avoided the distraction of polished but wrong explanations, it also explained its choices clearly. Bigger models, more parameters, more cost, yet worse results. If you rely on a single “trusted” LLM as a judge, you’re exposed to systematic bias. Subtle prompt design can swing outcomes. Alignment matters more than parameter count. And stacking multiple evaluators, not just scaling one, is the safer path. At Arato, we help teams plan, test, and run evaluation pipelines that reflect this reality, integrating diverse LLM judges, tracking bias, and ensuring results hold up in production.

256

4 Comments

Can clinical AI summaries be not just useful... but fun? We built something unusual at Abstractive Health: a free simulation game called Clinical Time Machine. Try for free here -> https://lnkd.in/epeuph9m Why a game? Because less than 1% of doctors have ever seen an actual AI summary of a real patient. Not because they don’t want to - it’s an access problem. Our summaries are really good on real patients - but it takes time to verify new users are doctors. And synthetic data doesn’t cut it. So we thought: what if we used famous historical patients instead? The first case we built is King George III - whose condition was long labeled as "madness." No spoilers if you know what it really was... We tracked down his original handwritten records from 200+ years ago, ran OCR on them, and fed them through our AI summary engine. Then we layered in RAG and agent-based reasoning to build a playable simulation. Doctors can now interact with the case like it’s their own: ask clinical questions, submit a differential, choose treatments, and watch the outcome. It’s built on top of our real product, and early testers said: “This made the board exams fun.” Hence Abstractive decided to build the first clinical AI game with chart summaries 🙂  Shout out to Caroline Reiner for the original idea! #HealthcareAI #ClinicalSimulation #FutureOfMedicine Doctors can diagnose King George III with our AI:

104

8 Comments

📣 Introducing RRE-1: A New Model for Retrieval Evaluation & Reranking for RAG based applications. Retrieval-Augmented Generation (RAG) has massively helped transform LLM-powered applications, but let’s face it—retrieval quality is still the weakest link in many pipelines. Poor retrieval is one of the major contributors of lower accuracy of your LLM applications. 🧐 That’s why we at AIMon built RRE-1, a powerful retrieval evaluation and reranker model designed to optimize ranking quality at scale. 🎯 👉 Why does this matter? Traditional retrieval methods like vector search are powerful, but they have limitations in effectiveness, especially when handling complex queries and domain specific knowledge. RRE-1 addresses this by offering: ✅ Offline evaluation of retrieval effectiveness, helping teams benchmark and refine their pipelines. ✅ Real-time low-latency reranking to ensure the most relevant context reaches the LLM. 👉 Who is this for? If you’re working on LLM agents, enterprise RAG systems, or AI copilots, this is for you. Our model ensures your retrieval pipeline is as optimized for your generative model—because great responses start with great retrieval. 🔎 Want to see it in action? Read more about AIMon RRE-1 in our blog post (link in comments). If you’re tackling retrieval challenges in your AI stack, we would love to chat! #AI #RAG #Retrieval #LLMs #MachineLearning

67

4 Comments

Nuanced is a structured context layer for AI coding tools: call graphs, symbol maps, and static analysis that make LLMs reliable. We just launched our explainer video: 📽️ https://lnkd.in/eA8wmYeA If you’ve ever banged your head against Claude Code or Cursor, you know the pain: LLMs can write code, but they don’t understand it. They hallucinate helpers, invent imports, and break prod, because they can’t trace real structure across your repo. Nuanced gives them that structure. No more guesswork. Just deterministic edits, grounded in your codebase.

47

6 Comments

🚗 LLMs vs Agents vs Agentic AI — Explained with LEGOs 🧱 Imagine you’re handed a pile of LEGO blocks and asked to build a car. Here’s how it plays out in the world of AI: 🔹 LLM (Large Language Model) The LLM is like a brilliant kid who has read every LEGO manual ever written. You ask, “How do I build a car?” and it gives you detailed instructions — maybe even alternatives based on what kind of car you want. But… it won’t actually build it for you. 🔹 Agent Now imagine giving those instructions to a helper (the agent) who can follow them. You tell the agent, “Here’s what the LLM said — now go assemble it.” The agent reads the blueprint, connects the pieces, and maybe even checks back if it’s unsure. It acts based on the LLM’s intelligence. 🔹 Agentic AI Agentic AI takes it one step further. You just say, “I need a car that can move fast and carry 2 passengers.” It breaks the problem down, figures out what kind of car to build, asks the LLM when needed, collects the right LEGO blocks, builds iteratively, tests it, improves it, and handles unexpected bumps along the way. In short, it thinks, plans, and acts — all autonomously. ⸻ 🤖 LLM = Knowledge 🧑‍🔧 Agent = Action 🧠 Agentic AI = Goal-driven Autonomy We’re entering a new era where AI doesn’t just answer — it accomplishes. #AgenticAI #LLMs #AIAgents #GenerativeAI #LLM #AIExplained #TechMetaphors #ArtificialIntelligence

56

3 Comments

2 Big things happened in AI last week - The first one could be highly disruptive, kind of like what NVDIA did to Intel. InceptionLabs released Mercury LLM. Traditional LLMs work by predicting the next word and thus work in serial, are slow and expensive. This uses Diffusion (which is already used for images but has struggled with text in the past).. potentially a fundamental change to the underlying LLM architecture.. it is faster and cheaper.. Second Manus.im (from China) released their AI agent that is a combination of Deep Research + Operator.. so it not only does the research but then takes action on it and gets stuff done... they managed to create a lot of buzz by creating scarcity.. even if it's a bit hyped up (supposedly runs on Anthropic, supposedly is a bit buggy) it is a view into the future. Which makes me question OpenAI's $20K agents.. competition and open source will be highly disruptive to their monetization strategy... what happens if Diffusion LLMs do take off.. Traditional LLM vendors will have lost a lot of money on an architecture that could be made obsolete before they could monetize it.. links in comments

36

7 Comments

A lot of people are reading "AI isn't replacing radiologists" and saying "AI isn't that good". But that's not the real story. Models like CheXNet could outperform human radiologists on some benchmarks as far back as 2017. Since then, hundreds of FDA-cleared tools have proven they can detect disease faster and more accurately than humans in some controlled tests. So why hasn't AI transformed radiology? Because the system won't let it. - The FDA requires new approval every time a model is retrained. - Malpractice insurers exclude liability for AI-only diagnoses (rightly so!) but often create gray zones for assistive AI. I think the lack of progress is that guardrails like these (and others) have prevented AI from reaching its potential: not to replace doctors, but to amplify them. I want to be clear about this: No radiologist (or doctor) should ever lose a job to AI. But I don't think any ever will. There aren't enough doctors as it is. The goal with AI in medicine should NOT be substitution, but augmentation. AI should eliminate the administrative grind, surface critical cases, and automate the boring stuff, so doctors can spend more time on what only humans can do: care, context, communication. The lesson from radiology isn't that AI failed. It's that AI needs the right incentives and freedom to grow, alongside human experts, not behind regulatory walls. Let's build a future where AI works with doctors, because our doctors need every beneficial tool we can build for them in order to meet the world's growing healthcare demand. https://lnkd.in/eWUyjexX

54

15 Comments

🔥 Just released MicroTuner: Fine-tune small LLMs for YOUR use case in minutes, not hours! ⚡️ 5x faster training with Unsloth AI ☁️ Serverless deployment with Modal 💰 Works with smaller models (1-7B params) 🤗 Automatic Hugging Face publishing

52

1 Comment

From an open-research point of view, probably the greatest thing about DeepSeek–R1 is how its RL training technique appears so straightforward and simple in comparison to the cumbersome approaches people were starting to think necessary for learning reasoning like Process Reward Models or Monte Carlo Tree Search. Once again it's party-time for the AI community which just discovered a whole new land of possibilities for building smarter AI models, and distillation in smaller models is one of the most exciting of these new promises.

410

16 Comments