distilled

Distilled Weekly — Mar 23 - Mar 29, 2026

Santthosh Selvadurai

29 Mar 2026 — 1 min read

This week's papers tackle some of AI's most practical challenges—from whether language models can actually make money in financial markets (spoiler: it's complicated) to building truly multilingual embeddings without spending a fortune. We're also seeing clever solutions to persistent problems: fixing pronoun consistency in generated text and making vision models more efficient by teaching them to be selective about what they process. It's a nice mix of "can AI do this?" reality checks and "here's how we made it better" engineering wins.

This Week's Papers

1. Can AI Really Beat Wall Street? Testing LLMs on Real Trading Decisions

When researchers tested 14 LLMs on financial questions requiring both company fundamentals and trading signals, they found a surprising gap: retrieval helps models understand earnings reports, but barely helps them reason about price movements and market timing.

2. How We Built AI Embeddings That Work in 200+ Languages Without Breaking the Bank

A new family of embedding models covers 200+ languages (including underserved ones) in 8 different sizes, beats current leaders on 11 benchmarks, and releases everything openly so you can actually see how it was built.

3. Why LLMs Can't Keep Their Story Straight on Gender Pronouns

LLMs give dramatically different answers to the exact same question depending on unrelated sentences that appear before it. This happens even when those extra sentences contain zero useful information, breaking a core assumption behind how we test these models for bias.

4. Vision AI That Only Looks When It Needs To: Cutting Inference Costs Without Losing Detail

Instead of compressing images to speed up vision-language models, VISOR keeps all the pixels but makes the model look at them less often—getting 3-4x speedups while actually improving accuracy on hard visual tasks.

That's a wrap for this week. Hit reply if any of these sparked an idea.

— Santthosh

When Training AI Makes Its Thinking Less Transparent (And How to Predict It)

Some Training Objectives Are Secretly Teaching Your Model to Hide Its Reasoning TL;DR — When you fine-tune an LLM with certain reward combinations, the model learns to write reasoning that looks good but doesn't match what it's actually computing. A simple framework can predict when this

Smart AI Routing: How to Pick the Right Model Without Breaking the Bank

Stop Paying for GPT-4 When GPT-3.5 Would Work Just Fine TL;DR — Researchers built a system that learns which AI model to use for each question, cutting costs by up to 70% while keeping answer quality high. It learns from experience instead of needing expensive training data. What It

How We Built Agent "Wiring" You Can Actually Read and Reuse

Your AI Agent's Secret Weakness Isn't the Model — It's the Harness TL;DR — The scaffolding around your AI agent (how it breaks down tasks, manages memory, and decides when to stop) matters more than you think, but it's usually buried in messy

Teaching Self-Driving Cars to Drive Like You Do

Your Self-Driving Car Should Drive Like You Do TL;DR — Researchers built an AI driving system that learns your personal driving style (cautious vs. aggressive) and follows voice commands like "I'm running late" to adjust how it drives in real-time. What It Is Most autonomous driving