Transformers exhibit inherent succinctness in their representations, enabling efficient processing of information. This research has been recognized as one of three outstanding papers for ICLR 2026.

Claude processes input words as numerical activations, which represent its internal thoughts. These activations are challenging to decode, making it difficult to interpret Claude's internal processing.

SHRDLU is an early natural-language understanding program developed by Terry Winograd at MIT between 1968 and 1970. It was written in Micro Planner and Lisp and ran on the PDP-6 operating system.

A Burla cluster processed 275 GB of Amazon reviews using 1,000 CPUs, categorizing them by various expressions of extreme sentiment. The system can display raw content filtered for profanity, rants, and other intense reactions.

GLiNER2 is a unified model for Named Entity Recognition, Text Classification, Structured Data Extraction, and Relation Extraction, consisting of 205 million parameters. It allows for efficient CPU-based inference without the need for complex pipelines or external API dependencies.

Large Language Models (LLMs) exhibit the ability to adopt different personas and behaviors based on social context. This flexibility is achieved through the presence of personality subnetworks within the models, rather than relying solely on external knowledge or prompting methods.

CAD Search allows users to find 3D models using natural language queries. The platform is powered by the ABC Dataset.