Sublime
An inspiration engine for ideas

🚨 Iterative Reasoning Preference Optimization 🚨
- Iterative algorithm for reasoning tasks: generate pairs & apply DPO+NLL
- Improves accuracy over iterations on GSM8K, MATH, ARC & beats baselines
E.g. Llama2-70B GSM8K: 55.6%->81.6% (88.7% maj32)
https://t.co/AGFLRAk5X3
🧵(1/5)... See more
If you can recall previous applicants, the optimal algorithm puts a twist on the familiar Look-Then-Leap Rule: a longer noncommittal period, and a fallback plan. For example, assume an immediate proposal
Brian Christian • Algorithms to Live By: The Computer Science of Human Decisions
I met Peter Thiel once and pitched him on this thing that automatically detects digital trends
He said that might be interesting in liquid markets, but in illiquid markets by the time a trend has arrived valuations are too high to make any real money
He offered to buy the tool anyways bc he... See more
goodalexanderx.comFrom Long Short-Term Memory (LSTMs) to New Gens and Digital Superforecasters
1/8 I used to build financial models with LSTMs many years ago, but they consistently fell short on finding meaningful patterns beyond basic "going up or down" predictions. Markets are far more complex than sequential models can capture.
reisearchx.com
New nature-published research brings AI model that simulates human decision-making and behavior with extreme accuracy
Helmholtz Munich researchers built Centaur, an AI that mirrors human choices and behaviors with striking accuracy by digesting millions of psychology-experiment... See more
- LLM Pro/Serious Use Comparison/Test: From 7B to 70B vs. ChatGPT! Winner: Synthia-70B-v1.2b
- LLM Chat/RP Comparison/Test: Dolphin-Mistral, Mistral-OpenOrca, Synthia 7B Winner: Mistral-7B-OpenOrca
- LLM Chat/RP Comparison/Test: Mistral 7B Base + Instruct
- LLM Chat/RP Comparison/Test (Euryale, FashionGPT, MXLewd, Synthia, Xwin) Winner: Xwin-LM-70B-V0.1
- New
