Related papers: Epistemic diversity across language models mitigat…
The rapid adoption of AI in Earth system science promises unprecedented speed and fidelity in the generation of climate information. However, this technological prowess rests on a fragile and unequal foundation: the current trajectory of AI…
Conversations transform individual knowledge into collective insight, enabling collaborators to solve problems more accurately than they could alone. Whether dialogues among large language models (LLMs) can replicate the synergistic gains…
Ideas generated by independent samples of humans tend to be more diverse than ideas generated from independent LLM samples, raising concerns that widespread reliance on LLMs could homogenize ideation and undermine innovation at a societal…
We provide an empirical investigation of the potential of pre-training vision-language models on an unprecedented scale: 100 billion examples. We find that model performance tends to saturate at this scale on many common Western-centric…
Human attention has become a scarce and strategically contested resource in digital environments. Content providers increasingly engage in excessive competition for visibility, often prioritizing attention-grabbing tactics over substantive…
Statistical averaging and asynchronous population dynamics as portfolio mechanisms are considered as the most important processes with which biodiversity contributes to ecosystem stability. However, portfolio theories usually regard…
In this paper, we investigate the relationship between diversity metrics, accuracy, and resiliency to natural image corruptions of Deep Learning (DL) image classifier ensembles. We investigate the potential of an attribution-based diversity…
Many of the kinds of language model used in speech understanding suffer from imperfect modeling of intra-sentential contextual influences. I argue that this problem can be addressed by clustering the sentences in a training corpus…
Post-trained language models produce less varied outputs than their base counterparts. This output diversity collapse undermines inference-time scaling methods that rely on varied samples, and risks homogenizing model outputs on creative…
Ecological systems are emergent features of ecological and adaptive dynamics of a community of interacting species. By natural selection through the abiotic environment and by co-adaptation within the community, species evolve, thereby…
Ecosystems are governed by dynamic processes such as competition for resources, reproduction and dispersal. These shape their biodiversity and how the system responds to change. Current approaches to modelling ecosystems, especially plants,…
Bilingual and multilingual language models offer a promising path toward scaling NLP systems across diverse languages and users. However, their performance often varies wildly between languages as prior works show that adding more languages…
Evolutionary algorithms have been successfully applied to a variety of optimisation problems in stationary environments. However, many real world optimisation problems are set in dynamic environments where the success criteria shifts…
Much of our understanding of ecological and evolutionary mechanisms derives from analysis of low-dimensional models: with few interacting species, or few axes defining "fitness". It is not always clear to what extent the intuition derived…
This paper presents a system for diversity-aware autonomous conversation leveraging the capabilities of large language models (LLMs). The system adapts to diverse populations and individuals, considering factors like background,…
Large language models increasingly function as epistemic agents -- entities that can 1) autonomously pursue epistemic goals and 2) actively shape our shared knowledge environment. They curate the information we receive, often supplanting…
Modern language model-based AI systems are remarkably powerful, yet their capabilities remain fundamentally capped by their human creators in three key ways. First, although a model's weights can be updated via fine-tuning, acquiring new…
Ensemble learning is a process by which multiple base learners are strategically generated and combined into one composite learner. There are two features that are essential to an ensemble's performance, the individual accuracies of the…
Universities are microcosms of urban ecosystems, with concentrated consumption patterns in food, transport, energy, and product usage. These environments not only contribute substantially to sustainability pressures but also provide a…
As the scope of machine learning broadens, we observe a recurring theme of algorithmic monoculture: the same systems, or systems that share components (e.g. training data), are deployed by multiple decision-makers. While sharing offers…