Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,341 papers · 148 categories

Trend · papers per month

71142213284 · Jun 202019922001200920182026
48 results for character inputs

Representation mixing combines character and phoneme inputs for flexible TTS synthesis.

problem Limited control over pronunciation in character or phoneme-based TTS systems.
method Representation mixing combines multiple linguistic inputs in a single encoder.
result Flexibility in choosing between character, phoneme, or mixed representations during inference.

Dynamic segmentation algorithm improves NMT performance by favoring character-level processing.

problem Suboptimal static segmentation choices in NMT systems.
method Adaptive Computation Time algorithm for dynamic segmentation, trainable end-to-end.
result The model prefers character-level processing when given the freedom to navigate different segmentation levels.

LAS neural network transcribes speech to characters without independence assumptions.

problem Transcribe speech to characters without making independence assumptions.
method Jointly learns listener and speller components; listener is a pyramidal recurrent network encoder; speller is an attention-based recurrent network decoder.
result Achieves 10.3% WER on Google voice search task subset, outperforming state-of-the-art CLDNN-HMM model.

Unsupervised model learns word and context embeddings from character sequences.

problem Learning meaningful word and context embeddings from unlabeled data.
method Character-aware neural architecture that jointly learns word and context embeddings.
result Compact encoders achieve high performance in downstream tasks.

Generative Neuro-Symbolic model learns from raw data with rich conceptual representations.

problem Learning rich, general-purpose conceptual representations from raw perceptual inputs.
method Generative Neuro-Symbolic (GNS) model combining symbolic and neural network approaches.
result Model learns from raw data and generalizes to 4 unique tasks.

End-to-end character-level model for text generation without delexicalization.

problem Generating text without delexicalization and tokenization.
method Character-level sequence-to-sequence model with attention mechanism, copy mechanism, and transfer learning.
result Competitive performance in text generation metrics.

Paper proposes BCNMCC for system identification with noisy input.

problem System identification with noisy input and impulsive output noise.
method Introduces BCV to NMCC algorithm to compensate for input noise bias.
result BCNMCC outperforms other algorithms in noisy input conditions.

Improves text-to-speech speed by interleaving character reading and audio synthesis.

problem Latency in text-to-speech models limits their use in time-sensitive tasks.
method Reinforcement learning to train an agent to choose the order of character reading and audio synthesis.
result The proposed method successfully balances latency and audio quality.

Sum-product networks enhance sequence modeling with higher-order factors.

problem Modeling complex relations in sequence data with first-order models.
method Combining sum-product networks with higher-order linear-chain conditional random fields.
result Improved performance in sequence labeling tasks compared to state-of-the-art methods.

Character scheme of small Seifert 3-manifolds is reduced if and only if no exceptional abelian character exists.

problem Character scheme of small Seifert 3-manifolds and its reduction condition.
method Complete description of the SL2(C)-character scheme of small Seifert 3-manifolds.
result Character scheme is reduced if and only if no exceptional abelian character exists, and exceptional abelian character has multiplicity 2.

Study on deformations of trivial character in SL_2(C) for groups.

problem Deforming the trivial character in SL_2(C) for groups.
method Analyzing maps satisfying the parallelogram identity on groups and relating them to deformations of the trivial character.
result The trivial character is a smooth point if and only if H_1(G,C) has dimension less than 2.

In this paper, we give a quantum interpretation of the Bismut-Chern character form (the loop space lifting of the Chern character form) as well as the Chern character form associated to a complex vector bundle with connection over a smooth manifold in the framework of supersymmetric quantum field theories developed by …

2007-11-24abs ↗pdf ↗

SmoothLLM defends LLMs from jailbreaking attacks by randomly perturbing inputs.

problem Adversaries can fool large language models into generating objectionable content.
method SmoothLLM randomly perturbs multiple copies of a prompt and aggregates predictions to detect adversarial inputs.
result SmoothLLM sets the state-of-the-art for robustness against various jailbreak attacks.

Study shows how certain knots and tori are detected by ideal points in character varieties.

problem Detecting essential twice-punctured tori in character varieties.
method Extending Tillmann's limiting character method to new families of knots and tori.
result Explicit determination of limiting characters at ideal points for specific knots and tori.

FinMem enhances LLM trading agents with layered memory and character design.

problem Developing purpose-driven LLM agents for financial decision-making.
method Integrates layered memory and character design modules into an LLM framework.
result Significantly enhanced trading performance in financial markets.

Deep neural nets predict click-through rates for sponsored search ads.

problem Predicting click-through rates for sponsored search ads.
method Two novel deep convolutional neural network approaches at character and word levels.
result Deep models significantly outperform baseline models and improve click-through rate prediction accuracy.

Computes dimensions of representation and character varieties for 2 and 3-dimensional orbifolds.

problem Computing dimensions of representation and character varieties for orbifolds.
method Analyzes varieties of representations and characters of hyperbolic and non-hyperbolic orbifolds, provides tools for computation.
result Shows that the dimension of a specific component of characters equals half the dimension of the variety of characters of the boundary.

Paper generates personalized fonts from a few characters.

problem Creating personalized fonts from a limited set of characters.
method Designs a network framework to extract and recombine character content and style using various neural networks.
result Generated characters are structurally similar to real characters.

Study of quantum decorated character stacks and their quantizations.

problem Quantization of decorated character stacks and their compatibility with cutting and gluing.
method Using stratified factorization homology, extend Fock and Goncharov's construction to include stacky points.
result Construction of categorical charts and flips on quantum decorated character stacks.

We study some basic properties of the variety of characters in PSL(2,C) of a finitely generated group. In particular we give an interpretation of its points as characters of representations. We construct 3-manifolds whose variety of characters has arbitrarily many components that do not lift to SL(2,C). We also study t…

2003-02-07abs ↗pdf ↗

Study character varieties of hyperbolic 3-manifolds using bundle methods.

problem Character varieties of hyperbolic 3-manifolds in once-punctured torus bundles.
method Restrict characters to the fibre and analyze branched covering maps.
result Infinite family of hyperbolic once-punctured bundles with unbounded genus.

Study the relationship between knot complements and their branched covers using abelian cord rings and SL(2,C) characters.

problem Understanding the relationship between knot complements and their branched covers.
method Utilizing Ng's abelian cord ring and SL(2,C) characters of the two-fold branched cover, focusing on augmentations and their ranks.
result Established a correspondence between trace-free SL(2,C) characters of a knot complement and augmentations, providing effective tools for finding representations.

mLSTM improves sequence modeling with better autoregressive density estimation.

problem Improving autoregressive density estimation in sequence modeling.
method Introduces mLSTM, a recurrent neural network combining LSTM and multiplicative recurrent networks.
result mLSTM outperforms standard LSTM and its variants in character-level language modeling tasks.

Researchers describe character varieties for Hopf links, proving geometric properties.

problem Character variety geometry of Hopf links with nn twists.
method Geometric descriptions of irreducible and totally reducible representations.
result Complete geometric description of SU(2)\mathrm{SU}(2)-character variety for r=2r=2.

Study of SU(2,1) character varieties on one-holed torus.

problem Characterize representations of mapping class group on SU(2,1) character variety.
method Explicit description of SU(2,1) character variety, use of Farey graph adaptation, and mapping class group action analysis.
result Description of an open domain of discontinuity for mapping class group action.