Letters of recommendation are a common tool used in graduate admissions. Most admissions systems require three letters for each applicant, burdening both letter writers and admissions committees with a heavy work load that may not be time well-spent. Most applicants do not have three research advisors who can comment meaningfully on research readiness, adding a large number of letters that are not useful. Ideally, letters of recommendation will showcase the students' promise for a research career, but in practice, the letters often do not fulfill this purpose. As a group of early and mid-career faculty who write dozens of letters every year for promising undergraduates, we are concerned and overburdened by the inefficiencies of the current system. In this open letter to the AAS Graduate Admissions Task Force, we offer an alternative to the current use of letters of recommendation: a portfolio submitted by the student, which highlights e.g., a paper, plot, or presentation that represents their past work and readiness for grad school, uploaded to a centralized system used by astronomy and astrophysics PhD programs. While we argue that we could eliminate letters in this new paradigm,
Acute poly-substance intoxication requires rapid, life-saving decisions under substantial uncertainty, as clinicians must rely on incomplete ingestion details and nonspecific symptoms. Effective diagnostic reasoning in this chaotic environment requires fusing unstructured, non-medical narratives (e.g. paramedic scene descriptions and unreliable patient self-reports or known histories), with structured medical data like vital signs. While Large Language Models (LLMs) show potential for processing such heterogeneous inputs, they struggle in this setting, often underperforming simple baselines that rely solely on patient histories. To address this, we present DeToxR (Decision-support for Toxicology with Reasoning), the first adaptation of Reinforcement Learning (RL) to emergency toxicology. We design a robust data-fusion engine for multi-label prediction across 14 substance classes based on an LLM finetuned with Group Relative Policy Optimization (GRPO). We optimize the model's reasoning directly using a clinical performance reward. By formulating a multi-label agreement metric as the reward signal, the model is explicitly penalized for missing co-ingested substances and hallucinating
Young adults often encounter challenges in career exploration. Self-guided interventions, such as the letter-exchange exercise, where participants envision and adopt the perspective of their future selves by exchanging letters with their envisioned future selves, can support career development. However, the broader adoption of such interventions may be limited without structured guidance. To address this, we integrated Large Language Model (LLM)-based agents that simulate participants' future selves into the letter-exchange exercise and evaluated their effectiveness. A one-week experiment (N=36) compared three conditions: (1) participants manually writing replies to themselves from the perspective of their future selves (baseline), (2) future-self agents generating letters to participants, and (3) future-self agents engaging in chat conversations with participants. Results indicated that exchanging letters with future-self agents enhanced participants' engagement during the exercise, while overall benefits of the intervention on future orientation, career self-concept, and psychological support remained comparable across conditions. We discuss design implications for AI-augmented i
It is well known that Charles Hermite kept an intense correspondence with many of the word's leading mathematicians of his time. This paper focuses on Hermite's letters to Francisco Gomes Teixeira, a Portuguese mathematician, who exchanged letters with Hermite for more than twenty years.
This paper presents a quasi-sequential optimal design framework for toxicology experiments, specifically applied to sea urchin embryos. The authors propose a novel approach combining robust optimal design with adaptive, stage-based testing to improve efficiency in toxicological studies, particularly where traditional uniform designs fall short. The methodology uses statistical models to refine dose levels across experimental phases, aiming for increased precision while reducing costs and complexity. Key components include selecting an initial design, iterative dose optimization based on preliminary results, and assessing various model fits to ensure robust, data-driven adjustments. Through case studies, we demonstrate improved statistical efficiency and adaptability in toxicology, with potential applications in other experimental domains.
Frequency of letters in a symbolic sequence ${\bf u}$ over a finite alphabet is one of the basic characteristics of ${\bf u}$. The notion of $k$-balancedness captures the property that the number of any letter occurring in two arbitrary factors of ${\bf u}$ of equal length differs at most by $k$. For a fixed integer $k$ and alphabet size $d\in \mathbb N$, we discuss possible frequencies of letters in $k$-balanced $d$-ary sequences. For the size $d$ of the alphabet, we introduce the notion of balancedness threshold $BT(d)$ and give an upper bound on it, where $BT(d)$ is the minimum $k$ such that there exists a $k$-balanced sequence over a $d$-letter alphabet for all possible letter frequencies.
We study the impact of generative AI on labor market signaling using the introduction of an AI-powered cover letter writing tool on a large online labor platform. Our data track both access to the tool and usage at the application level. Difference-in-differences estimates show that access to the tool increased textual alignment between cover letters and job posts and raised callback rates. Time spent editing AI-generated cover letter drafts is positively correlated with hiring success. After the tool's introduction, the correlation between cover letters' textual alignment and callbacks fell by 51%, consistent with what theory predicts if the AI technology reduces the signal content of cover letters. In response, employers shifted toward alternative signals, including workers' prior work histories.
Large language models (LLMs) struggle on simple tasks such as counting the number of occurrences of a letter in a word. In this paper, we investigate if ChatGPT can learn to count letters and propose an efficient solution.
Large Language Models (LLMs) have achieved unprecedented performance on many complex tasks, being able, for example, to answer questions on almost any topic. However, they struggle with other simple tasks, such as counting the occurrences of letters in a word, as illustrated by the inability of many LLMs to count the number of "r" letters in "strawberry". Several works have studied this problem and linked it to the tokenization used by LLMs, to the intrinsic limitations of the attention mechanism, or to the lack of character-level training data. In this paper, we conduct an experimental study to evaluate the relations between the LLM errors when counting letters with 1) the frequency of the word and its components in the training dataset and 2) the complexity of the counting operation. We present a comprehensive analysis of the errors of LLMs when counting letter occurrences by evaluating a representative group of models over a large number of words. The results show a number of consistent trends in the models evaluated: 1) models are capable of recognizing the letters but not counting them; 2) the frequency of the word and tokens in the word does not have a significant impact on t
Since clinical letters contain sensitive information, clinical-related datasets can not be widely applied in model training, medical research, and teaching. This work aims to generate reliable, various, and de-identified synthetic clinical letters. To achieve this goal, we explored different pre-trained language models (PLMs) for masking and generating text. After that, we worked on Bio\_ClinicalBERT, a high-performing model, and experimented with different masking strategies. Both qualitative and quantitative methods were used for evaluation. Additionally, a downstream task, Named Entity Recognition (NER), was also implemented to assess the usability of these synthetic letters. The results indicate that 1) encoder-only models outperform encoder-decoder models. 2) Among encoder-only models, those trained on general corpora perform comparably to those trained on clinical data when clinical information is preserved. 3) Additionally, preserving clinical entities and document structure better aligns with our objectives than simply fine-tuning the model. 4) Furthermore, different masking strategies can impact the quality of synthetic clinical letters. Masking stopwords has a positive im
In this paper we address the well-known problem of counting the number of $3n$-letter words that can be formed from a three-letter alphabet by decomposing it into four possible cases based on its remainder when divided by three. The solution to the problem also gives us some sums of trinomial coefficients.
Here we propose an extension of the (deterministic and the nondeterministic) finite automaton with translucent letters (DFAwtl and NFAwtl), which lies between these automata and their non-returning variants (that is, the nr-DFAwtl and the nr-NFAwtl). This new model works like a DFAwtl or an NFAwtl, but on seeing the end-of-tape marker, it may change its internal state and continue with its computation instead of just ending it, accepting or rejecting. This new type of automaton is called a repetitive deterministic or nondeterministic finite automaton with translucent letters (RDFAwtl or RNFAwtl). In the deterministic case, the new model is strictly more expressive than the DFAwtl, but less expressive than the nr-DFAwtl, while in the nondeterministic case, the new model is equivalent to the NFAwtl.
Deterministic and nondeterministic finite automata with translucent letters were introduced by Nagy and Otto more than a decade ago as Cooperative Distributed systems of a kind of stateless restarting automata with window size one. These finite state machines have a surprisingly large expressive power: all commutative semi-linear languages and all rational trace languages can be accepted by them including various not context-free languages. While the nondeterministic variant defines a language class with nice closure properties, the deterministic variant is weaker, however it contains all regular languages, some non-regular context-free languages, as the Dyck language, and also some languages that are not even context-free. In all those models for each state, the letters of the alphabet could be in one of the following categories: the automaton cannot see the letter (it is translucent), there is a transition defined on the letter (maybe more than one transitions in nondeterministic case) or none of the above categories (the automaton gets stuck by seeing this letter at the given state and this computation is not accepting). State-deterministic automata are recent models, where the
In toxicology research, experiments are often conducted to determine the effect of toxicant exposure on the behavior of mice, where mice are randomized to receive the toxicant or not. In particular, in fixed interval experiments, one provides a mouse reinforcers (e.g., a food pellet), contingent upon some action taken by the mouse (e.g., a press of a lever), but the reinforcers are only provided after fixed time intervals. Often, to analyze fixed interval experiments, one specifies and then estimates the conditional state-action distribution (e.g., using an ANOVA). This existing approach, which in the reinforcement learning framework would be called modeling the mouse's "behavioral policy," is sensitive to misspecification. It is likely that any model for the behavioral policy is misspecified; a mapping from a mouse's exposure to their actions can be highly complex. In this work, we avoid specifying the behavioral policy by instead learning the mouse's reward function. Specifying a reward function is as challenging as specifying a behavioral policy, but we propose a novel approach that incorporates knowledge of the optimal behavior, which is often known to the experimenter, to avoi
As the volume and complexity of nonclinical toxicology studies continue to increase, toxicologic pathology reporting faces persistent challenges, including fragmented sources of data (e.g., histopathology images, clinical pathology and other study data, adverse effects database, mechanistic literature), variable reporting timelines and heightened regulatory expectations. This white paper examines the emerging role of agentic artificial intelligence (AI) in addressing these issues through coordinated workflow orchestration, data integration, and pathologist-in-the-loop report generation. Based on a closed-door roundtable held during the 2025 Society of Toxicologic Pathology (STP) Annual Meeting and follow-on discussions, this paper synthesizes the perspectives of leading toxicologic pathologists, toxicologists, and AI developers. It outlines the key pain points in current reporting workflows, identifies realistic near-term use cases for agentic AI, and describes major adoption barriers including requirements for transparency, validation, and organizational readiness. A phased adoption roadmap and pilot design considerations are proposed to help support responsible evaluation and dep
Motivated by reformulating Yangian invariants in planar ${\cal N}=4$ SYM directly as $d\log$ forms on momentum-twistor space, we propose a purely algebraic problem of determining the arguments of the $d\log$'s, which we call "letters", for any Yangian invariant. These are functions of momentum twistors $Z$'s, given by the positive coordinates $α$'s of parametrizations of the matrix $C(α)$, evaluated on the support of polynomial equations $C(α) \cdot Z=0$. We provide evidence that the letters of Yangian invariants are related to the cluster algebra of Grassmannian $G(4,n)$, which is relevant for the symbol alphabet of $n$-point scattering amplitudes. For $n=6,7$, the collection of letters for all Yangian invariants contains the cluster ${\cal A}$ coordinates of $G(4,n)$. We determine algebraic letters of Yangian invariant associated with any "four-mass" box, which for $n=8$ reproduce the $18$ multiplicative-independent, algebraic symbol letters discovered recently for two-loop amplitudes.
We propose a geometrical approach to generate symbol letters of amplitudes/integrals in planar $\mathcal{N}=4$ Super Yang-Mills theory, known as {\it Schubert problems}. Beginning with one-loop integrals, we find that intersections of lines in momentum twistor space are always ordered on a given line, once the external kinematics $\mathbf{Z}$ is in the positive region $G_+(4,n)$. Remarkably, cross-ratios of these ordered intersections on a line, which are guaranteed to be positive now, nicely coincide with symbol letters of corresponding Feynman integrals, whose positivity is then concluded directly from such geometrical configurations. In particular, we reproduce from this approach the $18$ multiplicative independent algebraic letters for $n=8$ amplitudes up to three loops. Finally, we generalize the discussion to two-loop Schubert problems and, again from ordered points on a line, generate a new kind of algebraic letters which mix two distinct square roots together. They have been found recently in the alphabet of two-loop double-box integral with $n\geq9$, and they are expected to appear in amplitudes at $k+\ell\geq4$.
In celebration of the 2025 UN International Year of Quantum Science and Technology, this Resource Letter surveys the rapidly-growing field of scholarship in quantum information science and engineering (QISE) education. It is primarily written as a guide for educators wishing to get started teaching QISE using research-based teaching methods, as well as for discipline-based education research (DBER) practitioners looking to get started in this field. Topics covered include scoping the field of QISE education, research into student reasoning in QISE, research-based and research-inspired curricular materials from the high school to graduate level, research-based assessments, simulation and gamification tools, and tools for incorporating discussion of the societal and ethical implications of quantum technologies into the classroom.
Here we propose a variant of the nondeterministic finite automaton with translucent letters (NFAwtl) which, after reading and deleting a letter, does not return to the left end of its tape, but rather continues from the position of the letter just deleted. When the end-of-tape marker is reached, our automaton can decide whether to accept, to reject, or to continue, which means that it again reads the remaining tape contents from the beginning. This type of automaton, called a non-returning finite automaton with translucent letters or an nrNFAwtl, is strictly more expressive than the NFAwtl. We study the expressive capacity of this type of automaton and that of its deterministic variant. Also we are interested in closure properties of the resulting classes of languages and in decision problems.
We provide a systematic treatment of $D$-optimal design for binary regression and quantal response models in toxicology studies. For the two-parameter case, we provide an analytical equation (WC equation) for computing the $D$-optimal design quickly and when analytical solution is not available, we apply particle swarm optimization to solve for the $D$-optimal design. Examples with various link functions are given as well as the sensitivity functions. We extend the two-parameter case to three-parameter case by providing a neat formula for the determinant of the information matrix. We also suggest practitioners to work with the neat formula to derive optimal designs for three-parameter binary regression models.