Don't Pay Attention

The Transformer has become the de facto standard for large language models and a wide range of downstream tasks across various domains. Despite its numerous advantages like inherent training parallelism, the Transformer still faces key challenges due to its inability to effectively process sequences beyond a fixed context window and the quadratic complexity of its attention mechanism. These challenges have renewed interest in RNN-like architectures, which offer linear scaling with sequence length and improved handling of long-range dependencies, albeit with limited parallelism due to their inherently recurrent nature. In this paper, we propose Avey, a new neural foundational architecture that breaks away from both attention and recurrence. Avey comprises a ranker and an autoregressive neural processor, which collaboratively identify and contextualize only the most relevant tokens for any given token, regardless of their positions in the sequence. Specifically, Avey decouples sequence length from context width, thus enabling effective processing of arbitrarily long sequences. Experimental results show that Avey compares favorably to the Transformer across a variety of standard short-range NLP benchmarks, while notably excelling at capturing long-range dependencies.

Conversational Swarm Intelligence, a Pilot Study

Conversational Swarm Intelligence (CSI) is a new method for enabling large human groups to hold real-time networked conversations using a technique modeled on the dynamics of biological swarms. Through the novel use of conversational agents powered by Large Language Models (LLMs), the CSI structure simultaneously enables local dialog among small deliberative groups and global propagation of conversational content across a larger population. In this way, CSI combines the benefits of small-group deliberative reasoning and large-scale collective intelligence. In this pilot study, participants deliberating in conversational swarms (via text chat) (a) produced 30% more contributions (p<0.05) than participants deliberating in a standard centralized chat room and (b) demonstrated 7.2% less variance in contribution quantity. These results indicate that users contributed more content and participated more evenly when using the CSI structure.