An official website of the United States government
Theory is a critical component of the biological research process, and complements observational and experimental approaches. However, most biologists receive little training on how to frame a theoretical question and, thus, how to evaluate when theory has successfully answered the research question. Here, we develop a guide with six verbal framings for theoretical models in biology. These correspond to different personas one might adopt as a theorist: ‘Advocate’, ‘Explainer’, ‘Instigator’, ‘Mediator’, ‘Semantician' and ‘Tinkerer’. These personas are drawn from combinations of two starting points (pattern or mechanism) and three foci (novelty, robustness or conflict). We illustrate each of these framings with examples of specific theoretical questions, by drawing on recent theoretical papers in the fields of ecology and evolutionary biology. We show how the same research topic can be approached from slightly different perspectives, using different framings. We show how clarifying a model’s framing can debunk common misconceptions of theory: that simplifying assumptions are bad, more detail is always better, models show anything you want and modelling requires substantial maths knowledge. Finally, we provide a roadmap that researchers new to theoretical research can use to identify a framing to serve as a blueprint for their own theoretical research projects.
more »
« less
This content will become publicly available on July 27, 2026
Whose Boat Does it Float? Improving Personalization in Preference Tuning via Inferred User Personas
Language models are optimized to learn which responses you prefer, but they don't learn why you preferred a particular response. This limits their ability to tailor to personalized requests (e.g., "What should I eat for dinner? I'm vegetarian"), so we introduce a simple fix: have models infer personas that explain why users could prefer responses. We show training on these inferred personas leads to responses that are significantly more personalized for user needs.
more »
« less
- Award ID(s):
- 2403436
- PAR ID:
- 10608093
- Publisher / Repository:
- Association for Computational Linguistics
- Date Published:
- ISSN:
- 0736-587X
- Format(s):
- Medium: X
- Location:
- Vienna, Austria
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Have you eaten bread or rice recently? Or maybe something with oil in it? If you said yes, then you were likely eating foods made from seeds. Seeds are an important food source for humans and animals, and they make up our grains, lentils, nuts, and cooking oils. But seeds are also important for the plant itself, as they can be planted in the ground to grow new plants. In some plants, the seeds are covered with fruit. For example, the tomato fruit is full of seeds, and the avocado has a large seed inside. Other plants like pine trees have so-called naked seeds, with no fruit covering them. What exactly are seeds, and how do they make new plants? Why do seeds grow when you plant them and not in seed packets in the store? Keep reading and you will find these answers, and learn how fascinating, complex, and extraordinarily diverse seeds can be.more » « less
-
Abstract Organizations all over the world, both national and international, gather demographic data so that the progress of nations and peoples can be tracked. This data is often made available to the public in the form of aggregated national level data or individual responses (microdata). Product designers likewise conduct surveys to better understand their customer and create personas. Personas are archetypes of the individuals who will use, maintain, sell or otherwise be affected by the products created by designers. Personas help designers better understand the person the product is designed for. Unfortunately, the process of collecting customer information and creating personas is often a slow and expensive process. In this paper, we introduce a new method of creating personas, leveraging publicly available databanks of both aggregated national level and information on individuals in the population. A computational persona generator is introduced that creates a population of personas that mirrors a real population in terms of size and statistics. Realistic individual personas are filtered from this population for use in product development.more » « less
-
Personalized response selection systems are generally grounded on persona. However, a correlation exists between persona and empathy, which these systems do not explore well. Also, when a contradictory or off-topic response is selected, faithfulness to the conversation context plunges. This paper attempts to address these issues by proposing a suite of fusion strategies that capture the interaction between persona, emotion, and entailment information of the utterances. Ablation studies on the Persona-Chat dataset show that incorporating emotion and entailment improves the accuracy of response selection. We combine our fusion strategies and concept-flow encoding to train a BERT-based model which outperforms the previous methods by margins larger than 2.3% on original personas and 1.9% on revised personas in terms of hits@1 (top-1 accuracy), achieving a new state-of-the-art performance on the Persona-Chat datasetmore » « less
-
A "quine" is a deterministic program that prints itself. In this essay, I will show you a "gauguine": a probabilistic program that infers itself. A gauguine is repeatedly asked to guess its own source code. Initially, its chances of guessing correctly are of course minuscule. But as the gauguine observes more and more of its own previous guesses, it detects patterns of behavior and gains information about its inner workings. This information allows it to bootstrap self-knowledge, and ultimately discover its own source code. We will discuss how—and why—we might write a gauguine, and what we stand to learn by constructing one.more » « less
