BAIR

MAY 20, 2022

We recently published the Berkeley Crossword Solver (BCS), the current state of the art for solving American-style crossword puzzles. The most recent version is the current top-performing system on crossword puzzles from The New York Times, achieving 99.7% Crosswords are challenging for humans and computers alike.

148

148

Marktechpost

APRIL 11, 2024

The method has shown remarkable efficacy, achieving up to 30% improvement in story outline interestingness, 16% in crossword puzzle success, and 10% in concept coverage for generative tasks. Integrating Monte Carlo Tree Search with revision mechanisms allows for iterative refinement of outputs, enhancing decision-making processes.

Large Language Models 110

Large Language Models Artificial Intelligence Artificial Intelligence Data Analysis 110

TheSequence

JANUARY 14, 2024

The ML Behind the New York Times Crossword The New York Times shares some details of the hardwriting recognition models behind its famous crossworld puzzle —> Read more. . 🛠 Real World ML Ad Optimization at Pinterest Pinterest discusses the ML architecture powering its ad optimization infrastructure —> Read more.

Generative AI 105

Generative AI LLM AI AI 105

Towards AI

JULY 20, 2023

And it’s what we use when we do battle with a New York Times crossword puzzle, which requires different uses of word play, world knowledge and other linguistic artifacts. Because it requires models to have n-order logic. You can think of n-order logic as the Russian doll of logic, which is, logic nested within logic. accuracy ?

NLP 81

NLP Deep Learning Data Analysis Natural Language Processing 81

Topbots

DECEMBER 5, 2023

Empirical experiments demonstrate ToT’s effectiveness on challenging tasks such as Game of 24, Creative Writing, and Crosswords. The inspiration for ToT comes from “dual process” models in human decision-making, where fast, automatic decisions (System 1) are complemented by slower, deliberate decisions (System 2).

LLM 110

LLM Large Language Models Natural Language Processing Conversational AI 110

Topbots

APRIL 29, 2024

The authors also propose three new problems, Game of 24 , Creative Writing , and Crosswords , that require deductive, mathematical, commonsense, and lexical reasoning abilities. This integration facilitates a structured exploration of the tree of thoughts, incorporating both forward planning and the option to backtrack as necessary.

LLM 59

LLM Large Language Models Natural Language Processing AI Researcher 59

AssemblyAI

DECEMBER 2, 2022

And can you kind of exercise your neurons into better memory through crossword puzzles or deeper relationships or anything like that? It is something that you can hopefully address again, be involved in your brain health. - Can diet help us to avoid memory loss? Exercise the diet. Sleep and stress and learning new things is also helpful.

AI Modeling 130

AI Modeling AI AI Python 130