Searcharxiv⌕ Search

arXiv subjects

Anna Ivanova

Publications and source records attributed to Anna Ivanova.

12 recordsLinked to original sources
arXiv2026

DRIADA: A Python Toolkit for Cross-Scale Analysis of Single-Neuron Selectivity and Population Dynamics

Brain activity spans single-neuron, population, and network levels, and core questions in neural coding require moving between them. Yet current tools target a single paradigm and incompatible data formats, leaving cross-level questions hard to address. We present DRIADA, an open-source Python framework that unifies neural signals and time-aligned behavior in a shared data model, so selectivity testing, dimensionality reduction, and network analysis operate within a unified workflow. We evaluate it on synthetic data with known ground truth, hippocampal calcium imaging from 13~mice in an open field, and a simulated toroidal attractor network. In the hippocampal data, selectivity-based filtering restored a two-dimensional spatial embedding from a collapsed all-neuron embedding, while reverse analysis showed that ${\sim}57\%$ of neurons informative about leading manifold dimensions were not selective to any of the 11 measured behavioral features. On the toroidal benchmark, four independent modules recovered the expected topology. DRIADA makes cross-scale analysis routine across calcium imaging, spike trains, and simulated networks.

q-bio.NC↗
arXiv2026

INTENSE: Detecting and disentangling neuronal selectivity in calcium imaging data

Neurons encode information about the environment through their activity. As animals explore the environment, neurons rapidly acquire selectivity for distinct features of the external world; characterizing how these selectivity patterns emerge, reorganize, and overlap is key to linking neural activity to behavior and cognition. Calcium imaging in freely behaving animals can record large neuronal populations, but quantifying neuron-behavior selectivity directly from continuous fluorescence is challenging because both signals are temporally autocorrelated and calcium kinetics introduce time lags. Here we present INTENSE (INformation-Theoretic Evaluation of Neuronal SElectivity), an open-source framework that uses mutual information to detect neuron-behavior associations from raw calcium fluorescence data. INTENSE controls false discoveries using circular-shift permutation testing that preserves temporal structure and optimizes temporal delays to account for indicator kinetics and prospective/retrospective encoding. To separate genuine mixed selectivity from associations driven by behavioral covariance, INTENSE applies conditional mutual information-based disentanglement. We validated INTENSE on synthetic datasets, demonstrating robust detection across diverse signal-to-noise ratios and reliability conditions, whereas methods lacking temporal controls show poor performance. Applied to CA1 miniscope recordings in mice freely exploring an open field, INTENSE reveals robust selectivity to multiple variables (place, head direction, object interaction, locomotion) and refines mixed-selectivity estimates by distinguishing redundant from genuinely multi-variable encoding. Together, INTENSE enables high-throughput, information-theoretic selectivity mapping with principled control of temporal structure and behavioral covariance, bridging large-scale recordings to circuit-level hypotheses.

q-bio.NC↗
arXiv2025

What does it mean to understand language?

Language understanding entails not just extracting the surface-level meaning of the linguistic input, but constructing rich mental models of the situation it describes. Here we propose that because processing within the brain's core language system is fundamentally limited, deeply understanding language requires exporting information from the language system to other brain regions that compute perceptual and motor representations, construct mental models, and store our world knowledge and autobiographical memories. We review the existing evidence for this hypothesis, and argue that recent progress in cognitive neuroscience provides both the conceptual foundation and the methods to directly test it, thus opening up a new strategy to reveal what it means, cognitively and neurally, to understand language.

cs.CL↗
arXiv2025

How Do LLMs Use Their Depth?

Growing evidence suggests that large language models do not use their depth uniformly, yet we still lack a fine-grained understanding of their layer-wise prediction dynamics. In this paper, we trace the intermediate representations of several open-weight models during inference and reveal a structured and nuanced use of depth. Specifically, we propose a "Guess-then-Refine" framework that explains how LLMs internally structure their computations to make predictions. We first show that the top-ranked predictions in early LLM layers are composed primarily of high-frequency tokens, which act as statistical guesses proposed by the model due to the lack of contextual information. As contextual information develops deeper into the model, these initial guesses get refined into contextually appropriate tokens. We then examine the dynamic usage of layer depth through three case studies. (i) Multiple-choice task analysis shows that the model identifies appropriate options within the first half of the model and finalizes the response in the latter half. (ii) Fact recall task analysis shows that in a multi-token answer, the first token requires more computational depth than the rest. (iii) Part-of-speech analysis shows that function words are, on average, the earliest to be predicted correctly. To validate our results, we supplement probe-based analyses with causal manipulations in the form of activation patching and early-exiting experiments. Together, our results provide a detailed view of depth usage in LLMs, shedding light on the layer-by-layer computations that underlie successful predictions and providing insights for future works to improve computational efficiency in transformer-based models.

cs.CL↗
arXiv2024

Multilevel Interpretability Of Artificial Neural Networks: Leveraging Framework And Methods From Neuroscience

As deep learning systems are scaled up to many billions of parameters, relating their internal structure to external behaviors becomes very challenging. Although daunting, this problem is not new: Neuroscientists and cognitive scientists have accumulated decades of experience analyzing a particularly complex system - the brain. In this work, we argue that interpreting both biological and artificial neural systems requires analyzing those systems at multiple levels of analysis, with different analytic tools for each level. We first lay out a joint grand challenge among scientists who study the brain and who study artificial neural networks: understanding how distributed neural mechanisms give rise to complex cognition and behavior. We then present a series of analytical tools that can be used to analyze biological and artificial neural systems, organizing those tools according to Marr's three levels of analysis: computation/behavior, algorithm/representation, and implementation. Overall, the multilevel interpretability framework provides a principled way to tackle neural system complexity; links structure, computation, and behavior; clarifies assumptions and research priorities at each level; and paves the way toward a unified effort for understanding intelligent systems, may they be biological or artificial.

cs.AI↗
arXiv2024

Language models align with human judgments on key grammatical constructions

Do large language models (LLMs) make human-like linguistic generalizations? Dentella et al. (2023) ("DGL") prompt several LLMs ("Is the following sentence grammatically correct in English?") to elicit grammaticality judgments of 80 English sentences, concluding that LLMs demonstrate a "yes-response bias" and a "failure to distinguish grammatical from ungrammatical sentences". We re-evaluate LLM performance using well-established practices and find that DGL's data in fact provide evidence for just how well LLMs capture human behaviors. Models not only achieve high accuracy overall, but also capture fine-grained variation in human linguistic judgments.

cs.CL↗
arXiv2023

A Better Way to Do Masked Language Model Scoring

Estimating the log-likelihood of a given sentence under an autoregressive language model is straightforward: one can simply apply the chain rule and sum the log-likelihood values for each successive token. However, for masked language models (MLMs), there is no direct way to estimate the log-likelihood of a sentence. To address this issue, Salazar et al. (2020) propose to estimate sentence pseudo-log-likelihood (PLL) scores, computed by successively masking each sentence token, retrieving its score using the rest of the sentence as context, and summing the resulting values. Here, we demonstrate that the original PLL method yields inflated scores for out-of-vocabulary words and propose an adapted metric, in which we mask not only the target token, but also all within-word tokens to the right of the target. We show that our adapted metric (PLL-word-l2r) outperforms both the original PLL metric and a PLL metric in which all within-word tokens are masked. In particular, it better satisfies theoretical desiderata and better correlates with scores from autoregressive models. Finally, we show that the choice of metric affects even tightly controlled, minimal pair evaluation benchmarks (such as BLiMP), underscoring the importance of selecting an appropriate scoring metric for evaluating MLM properties.

cs.CL↗
arXiv2018

Does the brain represent words? An evaluation of brain decoding studies of language understanding

Language decoding studies have identified word representations which can be used to predict brain activity in response to novel words and sentences (Anderson et al., 2016; Pereira et al., 2018). The unspoken assumption of these studies is that, during processing, linguistic information is transformed into some shared semantic space, and those semantic representations are then used for a variety of linguistic and non-linguistic tasks. We claim that current studies vastly underdetermine the content of these representations, the algorithms which the brain deploys to produce and consume them, and the computational tasks which they are designed to solve. We illustrate this indeterminacy with an extension of the sentence-decoding experiment of Pereira et al. (2018), showing how standard evaluations fail to distinguish between language processing models which deploy different mechanisms and which are optimized to solve very different tasks. We conclude by suggesting changes to the brain decoding paradigm which can support stronger claims of neural representation.

cs.CL↗
arXiv2018

On asymptotic normality in estimation after a group sequential trial

We prove that in many realistic cases, the ordinary sample mean after a group sequential trial is asymptotically normal if the maximal number of observations increases. We derive that it is often safe to use naive confidence intervals for the mean of the collected observations, based on the ordinary sample mean. Our theoretical findings are confirmed by a simulation study.

math.ST↗
arXiv2018

On the sample mean after a group sequential trial

A popular setting in medical statistics is a group sequential trial with independent and identically distributed normal outcomes, in which interim analyses of the sum of the outcomes are performed. Based on a prescribed stopping rule, one decides after each interim analysis whether the trial is stopped or continued. Consequently, the actual length of the study is a random variable. It is reported in the literature that the interim analyses may cause bias if one uses the ordinary sample mean to estimate the location parameter. For a generic stopping rule, which contains many classical stopping rules as a special case, explicit formulas for the expected length of the trial, the bias, and the mean squared error (MSE) are provided. It is deduced that, for a fixed number of interim analyses, the bias and the MSE converge to zero if the first interim analysis is performed not too early. In addition, optimal rates for this convergence are provided. Furthermore, under a regularity condition, asymptotic normality in total variation distance for the sample mean is established. A conclusion for naive confidence intervals based on the sample mean is derived. It is also shown how the developed theory naturally fits in the broader framework of likelihood theory in a group sequential trial setting. A simulation study underpins the theoretical findings.

math.ST↗
arXiv2010

Identification of Convection Heat Transfer Coefficient of Secondary Cooling Zone of CCM based on Least Squares Method and Stochastic Approximation Method

The detailed mathematical model of heat and mass transfer of steel ingot of curvilinear continuous casting machine is proposed. The process of heat and mass transfer is described by nonlinear partial differential equations of parabolic type. Position of phase boundary is determined by Stefan conditions. The temperature of cooling water in mould channel is described by a special balance equation. Boundary conditions of secondary cooling zone include radiant and convective components of heat exchange and account for the complex mechanism of heat-conducting due to airmist cooling using compressed air and water. Convective heat-transfer coefficient of secondary cooling zone is unknown and considered as distributed parameter. To solve this problem the algorithm of initial adjustment of parameter and the algorithm of operative adjustment are developed.

math.OC↗