Neural attention has become central to many state-of-the-art models in natural language processing and related domains. Attention networks are an easy-to-train and effective method for softly simulating alignment; however, the approach does not marginalize over latent alignments in a probabilistic sense. This property …
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.
168,695 papers · 148 categories
Trend · papers per month
48 results for “Alignment Gain”
Inference-aware meta-alignment of LLMs reduces computational cost.
problem Aligning LLMs to diverse human preferences is challenging due to conflicting criteria.
method IAMA trains a base model to be aligned to multiple tasks via different inference-time alignment algorithms, using non-linear GRPO for optimization.
result IAMA enables effective alignment of LLMs to multiple criteria with limited computational budget.
Novel PO algorithms improve LLM alignment tasks.
problem Evaluating PO algorithms on LLM alignment is costly and noisy.
method Designed a diagnostic suite of MuJoCo tasks and datasets, used evolutionary strategies to discover specialized PO algorithms.
result Proposed MPO algorithms significantly outperform existing PO algorithms in LLM alignment.
Paper introduces a new metric to select optimal Graph Shift Operator for GNNs.
problem Empirical selection of Graph Shift Operator remains challenging.
method Introduces a novel alignment gain metric connecting geometric distortion to generalization bounds via spectral proxy.
result Provides a principled, computation-efficient criterion to rank and select optimal GSO.
Study improves stock return prediction by switching between economic states, outperforming traditional methods.
problem Improving stock return prediction across economic regimes.
method State-switching specification using the slope of the yield curve, with an Aligned Economic Index.
result The Aligned Economic Index outperforms traditional predictors, especially during market turbulence.
Unified framework simplifies DPO algorithms for LLM alignment.
problem Vast number of DPO variants complicates model alignment.
method Mutual information inspired unifying framework with flexible priors.
result Many DPO variants can be derived from the new framework.
PIPA aligns preferences without reinforcement learning, improving language model performance.
problem Aligning preferences in language models efficiently and without reinforcement learning.
method Formulates preference alignment as a Maximum Likelihood Estimation problem with prior constraints.
result PIPA algorithms achieve up to 10% performance improvement on benchmarks.
EXAGREE selects a stakeholder-aligned model to reduce conflicting explanations in machine learning.
problem Conflicting explanations from different attribution methods limit the adoption of machine learning models in safety-critical domains.
method EXAGREE is a two-stage framework that selects a Stakeholder-Aligned Explanation Model (SAEM) from a set of similar-performing models, maximizing Stakeholder-Machine Agreement (SMA).
result EXAGREE achieves simultaneous gains in faithfulness, plausibility, and fairness over baselines while preserving task accuracy.
Signed-permutation coordinate transport improves model alignment across checkpoints.
problem Improper alignment of coordinate-indexed objects across model checkpoints.
method Introduces sign-marginalized Hungarian matching and coordinate-preserving transport.
result Recovering signed-permutation gauge improves coordinate alignment and model performance.
Study tests how U.S. equity prices align with global asset frequencies using financial variables.
problem Testing whether U.S. equity prices align with global asset frequencies using financial variables.
method Examines SPX and RUT gaps, uses OIS-based funding, volatility, trading-friction, financial-condition variables, and residual information.
result Gains in fit survive broad-dollar neutralization, alternative blocks, PCA, residualization, and nested horizon selection, supporting reduced-form P-Q alignment.
The ability of a human being to extrapolate previously gained knowledge to other domains inspired a new family of methods in machine learning called transfer learning. Transfer learning is often based on the assumption that objects in both target and source domains share some common feature and/or data space. In this p…
Paper proposes CLAIR for efficient LLM fine-tuning across clients.
problem Fine-tuning large language models (LLMs) efficiently and collaboratively.
method Federated LoRA fine-tuning with Collaborative Low-rank Alignment and Identifiable Recovery (CLAIR).
result CLAIR achieves better performance and contamination detection compared to local fine-tuning.
Self-distillation improves constrained language generation by aligning models with target distributions.
problem Sparse and uninformative reward signals in constrained generation settings.
method Iteratively refining the base model through self-distillation, incorporating learned twist functions and proposals.
result Substantial gains in generation quality through improved model alignment with target distributions.
The paper develops methods to reduce deployment risk under dynamic covariate shifts.
problem Reduction of deployment risk under dynamic covariate shifts.
method Time-domain Poincare inequality and Jacobian-velocity theorem to identify and control directional tangent energy.
result Drift-aligned tangent regularization (DTR) reduces risk volatility and directional gain in low-rank drift regimes.
Unified model integrates text and time series for financial forecasting.
problem Challenges in integrating complementary modalities for improved forecasting.
method Modality-specific experts and cross-modal alignment framework.
result State-of-the-art performance on financial forecasting task.
New approach categorizes objective functions for embodied agents.
problem Understanding how objectives relate to each other and discovering new objectives.
method Introducing Action Perception Divergence (APD) to categorize objective functions.
result Introduces a spectrum of objectives from narrow to general, explaining various unsupervised objectives.
Decision forests, including Random Forests and Gradient Boosting Trees, have recently demonstrated state-of-the-art performance in a variety of machine learning settings. Decision forests are typically ensembles of axis-aligned decision trees; that is, trees that split only along feature dimensions. In contrast, many r…
Framework improves CATE estimation by aligning active learning with causal objectives.
problem High cost of outcome measurements limits CATE estimation.
method Causal-EPIG framework, targeting unobservable causal quantities.
result Strategies outperform standard baselines, revealing context-dependent optimal approaches.
Optimal feature learning strength improves generalization in deep networks.
problem Understanding how feature learning strength affects generalization in practical settings.
method Empirical studies and theoretical analysis of gradient flow dynamics in two-layer ReLU nets.
result Optimal feature learning strength yields substantial generalization gains, contrary to the prevailing intuition.
Spectral Gradient Descent Mitigates Anisotropy-Driven Misalignment: A Case Study in Phase Retrievalstat.ML
SpecGD mitigates misalignment in phase retrieval models with anisotropic inputs.
problem Misalignment during gradient descent in phase retrieval models with anisotropic inputs.
method Spectral gradient descent modifies gradient updates to preserve directional information and remove spike amplification.
result SpecGD removes spike amplification, leading to stable alignment and accelerated noise contraction.
Spectral feature learning improves IV regression for causal effect estimation.
problem Estimating causal effects in the presence of hidden confounders.
method Two-stage least squares estimator based on spectral features.
result Performance of the method depends on strong spectral alignment and slow eigenvalue decay.
This paper explores two techniques to improve the performance of text-dependent speaker verification systems based on deep neural networks. Firstly, we propose a general alignment mechanism to keep the temporal structure of each phrase and obtain a supervector with the speaker and phrase information, since both are rel…
Develops algorithm to find subgroups with different treatment effects in HIV patients.
problem Estimating treatment effects in EHR data with challenges like time-varying confounding.
method SDLD algorithm combining generalized interaction tree and longitudinal targeted maximum likelihood estimation.
result Identifies subgroups of HIV patients at higher risk of weight gain with dolutegravir-containing ARTs.
DLM-One speeds up language generation by 500x with continuous models.
problem Efficiently generating text sequences in natural language processing.
method Score-distillation of continuous diffusion language models.
result Achieves up to 500x speedup in inference time with competitive performance.
This work introduces a new metric to assess the fidelity of surrogate models to the underlying data-generating signal.
problem The limitations of fidelity-based explanations in explainable AI.
method Introduces the linearity score to quantify the extent of a regression network's linear decodability.
result High-fidelity surrogates can underperform compared to simpler models and even linear baselines trained directly on the data.
The paper introduces Relative Bias to quantify LLM bias systematically.
problem Quantifying bias in LLMs is challenging due to ambiguity and rapid model emergence.
method Relative Bias framework using Embedding Transformation and LLM-as-a-Judge methodologies.
result The two scoring methods show strong alignment, providing a systematic approach.
Improved neural framework for scaling entropic MOT with significant computational gains.
problem High computational overhead in multimarginal optimal transport.
method Neural Entropic MOT (NEMOT) using mini-batch training to reduce complexity.
result Significant speedups and feasibility improvements for multimarginal data.
Goal-driven Bayesian Optimal Experimental Design for Robust Decision-Making Under Model Uncertaintycs.LG
GoBOED optimizes experiments for specific decision-making objectives, improving downstream outcomes.
problem Reducing parameter uncertainty does not always improve decision-making in critical settings.
method Combines variational posterior surrogate and differentiable convex decision layer for gradient-based design optimization.
result GoBOED identifies designs that better align with specific decision objectives and reveals wider optimal design windows.
TempBalance boosts model performance with low data.
problem Low-data training and fine-tuning in model alignment.
method Inspired by HT-SR theory, TempBalance balances training quality across layers.
result TempBalance improves model performance as data decreases.
GO-CBED optimizes experiments for specific causal queries, improving efficiency.
problem Efficiently infer causal relationships with limited resources.
method Goal-oriented Bayesian framework that maximizes expected information gain on user-specified causal quantities.
result GO-CBED outperforms existing methods in various causal tasks, especially with limited budgets.
Supply Chain Management often requires independent organizations to work together to achieve shared objectives. This collaboration is necessary when coordinated actions benefit the group more than the uncoordinated efforts of individual firms. Despite the commonly reported benefits that can be gained in close relations…
Graph matching consists of aligning the vertices of two unlabeled graphs in order to maximize the shared structure across networks; when the graphs are unipartite, this is commonly formulated as minimizing their edge disagreements. In this paper, we address the common setting in which one of the graphs to match is a bi…
Scientific discovery is limited by hypothesis redundancy, and hybrid methods can exploit non-local exploration.
problem Limitation of scientific discovery due to hypothesis redundancy.
method Hybrid discovery systems combining structured local search with LLM-generated non-local proposals.
result Hybrid methods can exploit non-local exploration when three geometric conditions co-occur.
This paper proposes a method to gain extra supervision via multi-task learning for multi-modal video question answering. Multi-modal video question answering is an important task that aims at the joint understanding of vision and language. However, establishing large scale dataset for multi-modal video question answeri…
This paper surveys scalable automated alignment methods for LLMs.
problem Scalability issues in traditional human-annotated alignment methods for LLMs.
method Categorizes and discusses various automated alignment methods.
result Emerging automated alignment methods are effective and scalable.
Conformal Alignment ensures trustworthy outputs from foundation models.
problem Ensuring outputs from foundation models align with human values in high-stakes tasks.
method A framework that trains an alignment predictor using reference data to select trustworthy outputs.
result Conformal Alignment accurately identifies trustworthy outputs via lightweight training over moderate reference data.
The literature on structured prediction for NLP describes a rich collection of distributions and algorithms over sequences, segmentations, alignments, and trees; however, these algorithms are difficult to utilize in deep learning frameworks. We introduce Torch-Struct, a library for structured prediction designed to tak…
The social media revolution has changed the way that brands interact with consumers. Instead of spending their advertising budget on interstate billboards, more and more companies are choosing to partner with so-called Internet "influencers" --- individuals who have gained a loyal following on online platforms for the …
Dropout schedules can be optimized to significantly reduce model test loss.
problem Improving model performance in neural networks.
method Developed a mean-field theory of dropout at the edge of chaos, proposing front-loaded dropout schedules.
result Front-loaded dropout schedules reduce test loss by 18-35% over constant dropout.
LPL optimizes embeddings to align local neighborhoods, improving cross-lingual word alignment.
problem Aligning embeddings across different datasets and languages.
method Locality Preserving Loss (LPL) optimizes model to project embeddings while maintaining local neighborhoods and aligning them.
result LPL-based alignment leads to better and consistent accuracy, especially in small training set settings.
f-GRPO and Beyond: Divergence-Based Reinforcement Learning Algorithms for General LLM Alignmentcs.LG
Extends reinforcement learning alignment to scalar rewards, improving math reasoning.
problem Designing reinforcement learning algorithms for general LLM alignment.
method Introduces f-GRPO and f-HAL, estimating f-divergences between reward-aligned and unaligned distributions.
result Improves math reasoning RLVR tasks and mitigates reward hacking.
The paper proposes a learning-theoretic perspective on representation alignment.
problem Understanding how AI models' representations become aligned as they scale.
method Reviewing and connecting different notions of alignment, focusing on stitching.
result Relating properties of stitching to kernel alignment of representations.
FreSh shifts model's initial frequency spectrum to match target signal, improving neural representation performance.
problem MLPs' low-frequency bias limits capturing high-frequency details accurately.
method FreSh selects embedding hyperparameters to align model's initial output spectrum with target signal's spectrum.
result FreSh improves performance across various neural representation methods and tasks with minimal computational overhead.
Survey of alignment techniques for large language models.
problem Ensuring large language models align with human values.
method Analysis of diverse alignment methods and training paradigms.
result Preference-based methods offer more flexibility for nuanced alignment.
The paper addresses rigid alignment of noisy patches, providing a polynomial time algorithm and convergence conditions.
problem Finding a rigid alignment of overlapping local views (patches) that minimizes alignment error in a noisy setting.
method Characterizes non-degeneracy based on kernel and positivity of a matrix, provides polynomial time algorithm for testing non-degeneracy, and uses Riemannian gradient descent for alignment.
result The algorithm converges locally linearly to a non-degenerate perfect alignment under certain conditions.
Study connects database alignment and planted matching using Gaussian features.
problem Identify matching between correlated user features in anonymized databases.
method Derived results for database alignment and planted matching, showing connections and thresholds.
result Performance thresholds for database alignment converge to planted matching when feature dimensionality is sufficiently high.
Online news recommender systems aim to address the information explosion of news and make personalized recommendation for users. In general, news language is highly condensed, full of knowledge entities and common sense. However, existing methods are unaware of such external knowledge and cannot fully discover latent k…
FairPOT balances fairness and AUC performance by selectively transforming risk scores.
problem Balancing fairness and AUC performance in high-stakes domains.
method FairPOT uses proportional optimal transport to selectively transform risk scores.
result FairPOT consistently improves fairness with minimal AUC degradation or even positive gains.