arXiv subjects

Srinivas Ravishankar

Publications and source records attributed to Srinivas Ravishankar.

13 recordsLinked to original sources

arXiv2026

Conserved Kinematic Representations enable Zero-Shot Decoding in Handwriting BCIs

While intracortical Brain-Computer Interfaces (iBCIs) that decode imagined handwriting have achieved high communication rates for Latin scripts, they rely on observing every character in the alphabet during training. This poses a challenge in scaling to logographic languages (e.g., Chinese, Japanese), where the character set exceeds thousands of classes. The limitation highlights a fundamental question in motor neuroscience: does the motor cortex represent handwriting through the composition of shared kinematic primitives, that can be exploited by decoders? We introduce a computational framework for aligning neural activity to imagined kinematics in large datasets, enabling the training of a zero-shot capable machine learning algorithm for decoding unseen characters. Our model achieves 64% hits@3 retrieval on unseen letters, suggesting that neural representations of kinematic strokes are robustly conserved across different character contexts. This study provides a framework for dissecting conserved neural dynamics in large-scale intracortical datasets and offers strong evidence for a compositional basis of complex motor control. It also establishes a new paradigm for open-vocabulary iBCI communication with minimal recalibration burden on the user, crucial to increasing adoption of neuroprosthetics in logographic languages.

q-bio.NC↗

arXiv2026

Handwriting decoding as a challenging motor task for EEG Foundation Models

Recent attempts at creating Foundation Models (FMs) for Electroencephalography (EEG) have achieved state-of-the-art performance on multiple tasks including Motor Imagery (MI). These MI tasks have typically involved coarse classification between imagined limb movements. However, the development of foundation models necessitates diverse datasets, both for pretraining and evaluating the progress of these models. In this work, we propose handwriting decoding as a challenging motor task for FMs. We show that several existing datasets are potentially confounded, and introduce a dataset that more rigorously evaluates models. On this dataset, we find that current FMs, despite showing SOTA performance in multiple MI datasets are outperformed by smaller task-specific models. We also highlight challenges specific to EEG-based handwriting decoding to inform future work. In our 4-letter classification task, we show that (a) Knowledge of movement-onset is crucial to reported decoding performance in prior works, with average performance across subjects dropping from $41.3\%$ to $32.4\%$. (b) Increasing test-time signal quality provides significant performance improvements ($45\%$ to $78\%$ in our best subject) compared to scaling training data with single-trial EEG. (c) While scaling training data steadily improves decoding performance, existing FMs do not outperform specialist models in handwriting decoding. We make our code available at https://anonymous.4open.science/r/EEG-Handwriting-BCI-DFCD/

cs.HC↗

arXiv2025

Perceptogram: Reconstructing Visual Percepts and Presumptive Electrode Preference from EEG

Visual neural decoding from EEG has improved significantly due to diffusion models that can reconstruct high-quality images from decoded latents. While recent works have focused on relatively complex architectures to achieve good reconstruction performance from EEG, less attention has been paid to the source of this information. We present a unified framework that not only enables image reconstruction from EEG using a simple linear decoder, but also isolates interpretable EEG feature maps that support visual perception. Unlike prior approaches that rely on deep, opaque models, our method leverages the inherent structure of CLIP embeddings to keep the mapping linear. We show that training a simple linear decoder from EEG to CLIP latent space, followed by a frozen pre-trained diffusion model, is sufficient to decode images with state-of-the-art reconstruction performance. Beyond reconstruction, Perceptogram enables the visualization of presumptive electrode preference and EEG patterns, revealing interpretable EEG feature maps that correspond to distinct visual attributes, such as semantic class, texture, and hue. We thus use our framework, Perceptogram, to probe EEG signals at various levels of the visual information hierarchy.

q-bio.NC↗

arXiv2025

Monkey Perceptogram: Reconstructing Visual Representation and Presumptive Neural Preference from Monkey Multi-electrode Arrays

Understanding how the primate brain transforms complex visual scenes into coherent perceptual experiences remains a central challenge in neuroscience. Here, we present a comprehensive framework for interpreting monkey visual processing by integrating encoding and decoding approaches applied to two large-scale spiking datasets recorded from macaque using THINGS images (THINGS macaque IT Dataset (TITD) and THINGS Ventral Stream Spiking Dataset (TVSD)). We leverage multi-electrode array recordings from the ventral visual stream--including V1, V4, and inferotemporal (IT) cortex--to investigate how distributed neural populations encode and represent visual information. Our approach employs linear models to decode spiking activity into multiple latent visual spaces (including CLIP and VDVAE embeddings) and reconstruct images using state-of-the-art generative models. We further utilize encoding models to map visual features back to neural activity, enabling visualization of the "preferred stimuli" that drive specific neural ensembles. Analyses of both datasets reveal that it is possible to reconstruct both low-level (e.g., color, texture) and high-level (e.g., semantic category) features of visual stimuli from population activity, with reconstructions preserving key perceptual attributes as quantified by feature-based similarity metrics. The spatiotemporal spike patterns reflect the ventral stream's hierarchical organization with anterior regions representing complex objects and categories. Functional clustering identifies feature-specific neural ensembles, with temporal dynamics show evolving feature selectivity post-stimulus. Our findings demonstrate feasible, generalizable perceptual reconstruction from large-scale monkey neural recordings, linking neural activity to perception.

q-bio.NC↗

arXiv2025

Decoding Imagined Handwriting from EEG

Patients with extreme forms of paralysis face challenges in communication, adversely impacting their quality of life. Recent studies have reported higher-than-chance performance in decoding handwritten letters from EEG signals, potentially allowing these subjects to communicate. However, all prior works have attempted to decode handwriting from EEG during actual motion. Furthermore, they assume that precise movement-onset is known. In this work, we focus on settings closer to real-world use where either movement onset is not known or movement does not occur at all, fully utilizing motor imagery. We show that several existing studies are affected by confounds that make them inapplicable to the imagined handwriting setting. We also investigate how sample complexity affects handwriting decoding performance, guiding future data collection efforts. Our work shows that (a) Sample complexity analysis in single-trial EEG reveals a noise ceiling, which can be alleviated by averaging over trials. (b) Knowledge of movement-onset is crucial to reported performance in prior works. (c) Fully imagined handwriting can be decoded from EEG with higher-than-chance performance. Taken together, these results highlight both the unique challenges and avenues to pursue to build a practical EEG-based handwriting BCI.

cs.HC↗

arXiv2023

Adaptive recurrent vision performs zero-shot computation scaling to unseen difficulty levels

Humans solving algorithmic (or) reasoning problems typically exhibit solution times that grow as a function of problem difficulty. Adaptive recurrent neural networks have been shown to exhibit this property for various language-processing tasks. However, little work has been performed to assess whether such adaptive computation can also enable vision models to extrapolate solutions beyond their training distribution's difficulty level, with prior work focusing on very simple tasks. In this study, we investigate a critical functional role of such adaptive processing using recurrent neural networks: to dynamically scale computational resources conditional on input requirements that allow for zero-shot generalization to novel difficulty levels not seen during training using two challenging visual reasoning tasks: PathFinder and Mazes. We combine convolutional recurrent neural networks (ConvRNNs) with a learnable halting mechanism based on Graves (2016). We explore various implementations of such adaptive ConvRNNs (AdRNNs) ranging from tying weights across layers to more sophisticated biologically inspired recurrent networks that possess lateral connections and gating. We show that 1) AdRNNs learn to dynamically halt processing early (or late) to solve easier (or harder) problems, 2) these RNNs zero-shot generalize to more difficult problem settings not shown during training by dynamically increasing the number of recurrent iterations at test time. Our study provides modeling evidence supporting the hypothesis that recurrent processing enables the functional advantage of adaptively allocating compute resources conditional on input requirements and hence allowing generalization to harder difficulty levels of a visual reasoning problem without training.

cs.CV↗

arXiv2022

CBR-iKB: A Case-Based Reasoning Approach for Question Answering over Incomplete Knowledge Bases

Knowledge bases (KBs) are often incomplete and constantly changing in practice. Yet, in many question answering applications coupled with knowledge bases, the sparse nature of KBs is often overlooked. To this end, we propose a case-based reasoning approach, CBR-iKB, for knowledge base question answering (KBQA) with incomplete-KB as our main focus. Our method ensembles decisions from multiple reasoning chains with a novel nonparametric reasoning algorithm. By design, CBR-iKB can seamlessly adapt to changes in KBs without any task-specific training or fine-tuning. Our method achieves 100% accuracy on MetaQA and establishes new state-of-the-art on multiple benchmarks. For instance, CBR-iKB achieves an accuracy of 70% on WebQSP under the incomplete-KB setting, outperforming the existing state-of-the-art method by 22.3%.

cs.CL↗

arXiv2022

A Benchmark for Generalizable and Interpretable Temporal Question Answering over Knowledge Bases

Knowledge Base Question Answering (KBQA) tasks that involve complex reasoning are emerging as an important research direction. However, most existing KBQA datasets focus primarily on generic multi-hop reasoning over explicit facts, largely ignoring other reasoning types such as temporal, spatial, and taxonomic reasoning. In this paper, we present a benchmark dataset for temporal reasoning, TempQA-WD, to encourage research in extending the present approaches to target a more challenging set of complex reasoning tasks. Specifically, our benchmark is a temporal question answering dataset with the following advantages: (a) it is based on Wikidata, which is the most frequently curated, openly available knowledge base, (b) it includes intermediate sparql queries to facilitate the evaluation of semantic parsing based approaches for KBQA, and (c) it generalizes to multiple knowledge bases: Freebase and Wikidata. The TempQA-WD dataset is available at https://github.com/IBM/tempqa-wd.

cs.CL↗

arXiv2021

A Two-Stage Approach towards Generalization in Knowledge Base Question Answering

Most existing approaches for Knowledge Base Question Answering (KBQA) focus on a specific underlying knowledge base either because of inherent assumptions in the approach, or because evaluating it on a different knowledge base requires non-trivial changes. However, many popular knowledge bases share similarities in their underlying schemas that can be leveraged to facilitate generalization across knowledge bases. To achieve this generalization, we introduce a KBQA framework based on a 2-stage architecture that explicitly separates semantic parsing from the knowledge base interaction, facilitating transfer learning across datasets and knowledge graphs. We show that pretraining on datasets with a different underlying knowledge base can nevertheless provide significant performance gains and reduce sample complexity. Our approach achieves comparable or state-of-the-art performance for LC-QuAD (DBpedia), WebQSP (Freebase), SimpleQuestions (Wikidata) and MetaQA (Wikimovies-KG).

cs.CL↗

arXiv2021

SYGMA: System for Generalizable Modular Question Answering OverKnowledge Bases

Knowledge Base Question Answering (KBQA) tasks that in-volve complex reasoning are emerging as an important re-search direction. However, most KBQA systems struggle withgeneralizability, particularly on two dimensions: (a) acrossmultiple reasoning types where both datasets and systems haveprimarily focused on multi-hop reasoning, and (b) across mul-tiple knowledge bases, where KBQA approaches are specif-ically tuned to a single knowledge base. In this paper, wepresent SYGMA, a modular approach facilitating general-izability across multiple knowledge bases and multiple rea-soning types. Specifically, SYGMA contains three high levelmodules: 1) KB-agnostic question understanding module thatis common across KBs 2) Rules to support additional reason-ing types and 3) KB-specific question mapping and answeringmodule to address the KB-specific aspects of the answer ex-traction. We demonstrate effectiveness of our system by evalu-ating on datasets belonging to two distinct knowledge bases,DBpedia and Wikidata. In addition, to demonstrate extensi-bility to additional reasoning types we evaluate on multi-hopreasoning datasets and a new Temporal KBQA benchmarkdataset on Wikidata, namedTempQA-WD1, introduced in thispaper. We show that our generalizable approach has bettercompetetive performance on multiple datasets on DBpediaand Wikidata that requires both multi-hop and temporal rea-soning

cs.CL↗

arXiv2021

Leveraging Abstract Meaning Representation for Knowledge Base Question Answering

Knowledge base question answering (KBQA)is an important task in Natural Language Processing. Existing approaches face significant challenges including complex question understanding, necessity for reasoning, and lack of large end-to-end training datasets. In this work, we propose Neuro-Symbolic Question Answering (NSQA), a modular KBQA system, that leverages (1) Abstract Meaning Representation (AMR) parses for task-independent question understanding; (2) a simple yet effective graph transformation approach to convert AMR parses into candidate logical queries that are aligned to the KB; (3) a pipeline-based approach which integrates multiple, reusable modules that are trained specifically for their individual tasks (semantic parser, entity andrelationship linkers, and neuro-symbolic reasoner) and do not require end-to-end training data. NSQA achieves state-of-the-art performance on two prominent KBQA datasets based on DBpedia (QALD-9 and LC-QuAD1.0). Furthermore, our analysis emphasizes that AMR is a powerful tool for KBQA systems.

cs.CL↗

arXiv2020

Leveraging Semantic Parsing for Relation Linking over Knowledge Bases

Knowledgebase question answering systems are heavily dependent on relation extraction and linking modules. However, the task of extracting and linking relations from text to knowledgebases faces two primary challenges; the ambiguity of natural language and lack of training data. To overcome these challenges, we present SLING, a relation linking framework which leverages semantic parsing using Abstract Meaning Representation (AMR) and distant supervision. SLING integrates multiple relation linking approaches that capture complementary signals such as linguistic cues, rich semantic representation, and information from the knowledgebase. The experiments on relation linking using three KBQA datasets; QALD-7, QALD-9, and LC-QuAD 1.0 demonstrate that the proposed approach achieves state-of-the-art performance on all benchmarks.

cs.CL↗

arXiv2018

Revisiting Simple Neural Networks for Learning Representations of Knowledge Graphs

We address the problem of learning vector representations for entities and relations in Knowledge Graphs (KGs) for Knowledge Base Completion (KBC). This problem has received significant attention in the past few years and multiple methods have been proposed. Most of the existing methods in the literature use a predefined characteristic scoring function for evaluating the correctness of KG triples. These scoring functions distinguish correct triples (high score) from incorrect ones (low score). However, their performance vary across different datasets. In this work, we demonstrate that a simple neural network based score function can consistently achieve near start-of-the-art performance on multiple datasets. We also quantitatively demonstrate biases in standard benchmark datasets, and highlight the need to perform evaluation spanning various datasets.

cs.AI↗