Show HN: Trained an LLM to predict "What will Trump do?" (opens in new tab)

(huggingface.co)

10 pointsbturtel1mo ago2 comments

Hey HN! I RL-tuned an open-source LLM (gpt-oss-120b — 120B MoE, but only 5.1B active params) to predict "What will Trump do?" in any situation, trained on nothing but public news collected automatically from search queries. The trained model beats GPT-5, and both dataset and trained model are open sourced.

Data generation: Generated 2,108 binary forecasting questions from just a search query and a date range using the Lightning Rod SDK (https://github.com/lightning-rod-labs/lightningrod-python-sd...). Questions are generated from historic news articles — like "Will Trump impose 25% tariffs on Mexico by March 1?" — and resolved by checking what actually happened after the deadline. No human annotation — the whole pipeline is automated.

Training: GRPO with Brier score as the reward signal. LoRA rank 32, 50 training steps.

Results: Slight accuracy edge over GPT-5 (Brier 0.194 vs 0.200), but big gains in calibration — the RL-tuned model produces much better probabilities (ECE 0.079 vs 0.091).

Dataset: https://huggingface.co/datasets/LightningRodLabs/WWTD-2025

This is a fully automated way to spin up domain expert LLMs from public web data with just a few search queries, no labeling/annotation required.

I’d love any feedback, or suggestions for what domain expert to train next!

2 comments

sleno1mo ago

interesting...what were some examples of things trump did that your model got right and gpt-5 got wrong?

bturtelOP1mo ago

Great question! It's probabilistic so not really "right vs wrong" on any single question, but who better estimated the likelihood. One big difference shows up when there's no useful context - we ran the same eval WITHOUT including any useful up-to-date context with questions. In this case, GPT-5 stays overconfident and its BSS drops to -11.3% (vs -4.3% ours) - worse than just guessing the base rate. So one advantage of the RL training is just learning to know what you don't know, and identify when there's real signal.

j / k navigate · click thread line to collapse