New estimator corrects bias in CKA for sparsely sampled neurons.
problem Bias in CKA for sparsely sampled neurons.
method Novel estimator that corrects for input and feature sampling.
result Reliable model-to-brain alignment with sparsely sampled neurons.
In machine learning the best performance on a certain task is achieved by fully supervised methods when perfect ground truth labels are available. However, labels are often noisy, especially in remote sensing where manually curated public datasets are rare. We study the multi-modal cadaster map alignment problem for wh…
Differentiable sampling corrects alignment issues in neural machine translation.
problem Incorrect alignment of reference words and sampled output in scheduled sampling.
method Optimizes alignment probability based on model's soft alignment prediction.
result Improves BLEU score compared to maximum likelihood and scheduled sampling.
One-step Bellman alignment improves online RL by reducing task mismatch.
problem Online RL struggles with task similarity defined by rewards or transitions.
method One-step Bellman alignment and re-weighted targeting (RWT) to correct task mismatch.
result Regret bounds show task shift complexity, not target MDP, affects performance.
Develops methods to correct bias in AI feedback for more accurate alignment.
problem Systematic bias in AI feedback compared to human labels.
method Two debiased alignment methods: DDPO and DIPO.
result Methods improve alignment efficiency and performance close to human-labeled data.
We propose a novel framework for combining datasets via alignment of their intrinsic geometry. This alignment can be used to fuse data originating from disparate modalities, or to correct batch effects while preserving intrinsic data structure. Importantly, we do not assume any pointwise correspondence between datasets…
Theoretical analysis shows LLMs can self-correct responses through in-context learning.
problem Understanding how large language models improve through self-correction.
method Theoretical analysis based on simplified alignment task, focusing on softmax attention, multi-head attention, and MLP blocks.
result LLMs can refine responses in an in-context way when given accurate self-examinations as rewards.
LLMs struggle with zero-shot annotation tasks due to model-internalized priors.
problem Impact of model-internalized priors on LLM performance in zero-shot annotation tasks.
method Investigated three dimensions: familiarity, decision stickiness, and susceptibility to misaligned task definitions.
result Nearly two-thirds of zero-shot errors are resistant to correction, with a rescue rate of 34.8%. Definition-Specific Familiarity (DSF) shows a positive association with model performance.
Generative model improves wind field downscaling from coarse climate models.
problem Limited spatial resolution and biases in GCMs for wind energy studies.
method SerpentFlow for domain alignment and conditional fine-scale learning.
result Improved spatial coherence, inter-variable consistency, robustness under climate change.
Bayesian method corrects for model selection multiplicity in regression.
problem Model selection multiplicity in regression analysis.
method Developed a Bayesian prior distribution based on Holm procedure analogy.
result Adequate multiplicity correction requires sparsity not provided by recommended priors.
An efficient algorithm for aligning diffusion trees to networks with information asymmetry.
problem Aligning diffusion trees to networks with information asymmetry.
method Tree correlation tests for extracting alignment information.
result Explicit lower bounds on the probability of correct matches for each vertex on the diffusion tree.
A new protocol corrects confounding effects to measure alignment-induced activation shifts accurately.
problem Confounding effects in measuring alignment-induced activation shifts using naive methods.
method Introduces a four-variant decomposition to separate alignment shift from template effects.
result Correctly measures alignment-induced activation shifts, recovering behaviorally active subspace.
Partial soft-matching distance improves neural representation comparison by allowing some neurons to remain unmatched.
problem Neural representations are noisy and contain outliers, making traditional matching methods unreliable.
method Extends soft-matching distance to a partial optimal transport setting, allowing some neurons to remain unmatched.
result Partial soft-matching provides robust correspondences that are more reliable under noise and outliers.
Sequence probability predicts correctness in LLMs, but not for repeated prompts
problem Predicting correctness in large language models
method Quantifying sequence probability and correctness across different levels
result Higher sequence probability often predicts correctness across prompt-answer pairs
Sig-PCA integrates model outputs and observations to correct model biases.
problem Improving model accuracy and reliability by correcting biases and numerical approximations.
method Sig-PCA framework that combines summary statistics from model outputs with localized observations via a neural network.
result Corrects model outputs to align closely with observational data, preserving essential statistical information.
This paper improves model fusion by training-time neuron alignment, reducing barriers in multi-model fusion.
problem Diverse neuron permutations across different settings hinder model fusion performances.
method Training-time neuron alignment using fixed neuron anchors to reduce training-time permutations.
result Training-time neuron alignment improves fusion of pretrained models and federated learning performances.
We use automatic speech recognition to assess spoken English learner pronunciation based on the authentic intelligibility of the learners' spoken responses determined from support vector machine (SVM) classifier or deep learning neural network model predictions of transcription correctness. Using numeric features produ…
In the context of machine learning, disparate impact refers to a form of systematic discrimination whereby the output distribution of a model depends on the value of a sensitive attribute (e.g., race or gender). In this paper, we propose an information-theoretic framework to analyze the disparate impact of a binary cla…
FJS method improves multinomial classification accuracy.
problem Improving multinomial classification accuracy under dataset shift.
method Derive FJS representation and propose alternative methods.
result Factorizable joint shift is not fully identifiable without additional assumptions.
GICDM corrects hubness in embedding spaces for better generative model evaluation.
problem Hubness phenomenon distorts distances in high-dimensional embedding spaces.
method Generative ICDM (GICDM) using multi-scale extension to correct neighborhood estimation.
result GICDM resolves hubness-induced failures and improves metric behavior.
New algorithm for robust circular coordinates in recurrent time series data.
problem Inefficient and sensitive methods for finding circular coordinates on recurrent data.
method Subsampling, aligning, and averaging to correct uneven sampling density.
result More robust and efficient circular coordinates for neuronal recordings.
SRPO improves AI alignment with human preferences through self-improvement and task-independent optimization.
problem AI models trained with RLHF lack self-correction mechanisms and struggle with task generalization.
method SRPO formulates the preference learning problem as a min-max objective, optimizing a self-improvement policy and a generative policy in an adversarial fashion, making the solution task-independent.
result SRPO outperforms existing methods, achieving 90% AI Win-Rate on XSum and 56% on Arena-Hard prompts after a single revision.
Optimizes chip component placement with self-alignment for SMT technology.
problem Achieving precise component placement on PCB during SMT process.
method Proposed machine learning algorithms (SVR and RFR) to predict component positions and developed non-linear optimization model.
result RFR model outperforms in predicting component positions before reflow.
Partial recovery of node mappings between correlated graphs is possible under specific conditions.
problem Recovering a one-to-one mapping between nodes of two correlated graphs with a fraction of correct matches.
method Analyzing the graph isomorphism problem as a noisy version, considering Erdős-Rényi graphs, and providing conditions for partial recovery.
result Necessary and sufficient conditions for partial recovery of node mappings in correlated graphs are given.
Response time improves alignment with diverse human preferences.
problem Standard aggregation of feedback ignores heterogeneity and anonymity.
method Augmenting feedback with response time data and modeling decisions with DDM.
result Estimator of heterogeneous preferences converges to true average preference.
Two log-linear approximations speed up optimal transport for deep learning applications.
problem Computing optimal transport in high dimensions is computationally expensive.
method Locality-sensitive hashing (LSH) and Nyström approximation with LSH-based sparse corrections.
result Log-linear time algorithms for entropy-regularized OT perform well in high-dimensional spaces.
New method improves model calibration by adjusting confidence based on prediction correctness.
problem Improving model confidence alignment with true class probabilities.
method Post-hoc calibration objective using transformed samples for training.
result Competitive calibration performance on in-distribution and out-of-distribution test sets.
A new method improves data representation for diverse tasks.
problem Learning meaningful representations for tasks like batch correction and counterfactual inference.
method Contrastive Mixture of Posteriors (CoMP) method using misalignment penalties.
result CoMP achieves state-of-the-art performance on challenging tasks.
LLM evaluation suffers from systematic biases and lacks reliable positive judgments.
problem LLM evaluation suffers from systematic biases and lacks reliable positive judgments.
method Formulate LLM evaluation as a positive-unlabelled learning problem and propose a geometric auditing framework based on Partial Optimal Transport.
result Improved alignment with human preferences, increased robustness to presentation biases, and interpretable confidence estimates.
PolicySynth improves synthetic data alignment with real data for better campaign decisions.
problem Synthetic data used in decision support systems often leads to incorrect decisions.
method PolicySynth framework that conditions synthetic data on churn scorer to align with real data decisions.
result PolicySynth achieves high strategy simulation fidelity (0.923-0.960) on churn and acquisition datasets.
Fine-tuning LLMs with observational data can lead to spurious correlations, but DeconfoundLM can mitigate this.
problem Aligning LLMs with human preferences and business objectives using observational data.
method DeconfoundLM, a method that removes confounders from reward signals.
result DeconfoundLM improves recovery of causal relationships and mitigates spurious correlations.
Optimal transport aligns source and target distributions for domain adaptation.
problem Unsupervised domain adaptation with joint class-conditional and label shifts.
method Minimizes importance weighted loss and Wasserstein distance for aligned marginals and class-conditional distributions.
result Our method outperforms competitors on various domain adaptation tasks.
We explain how neural networks learn to solve modular addition tasks.
problem How two-layer neural networks learn to solve modular addition tasks.
method Formalized a diversification condition during training, proving it allows the network to approximate the correct logic for modular addition.
result Neural networks can robustly identify the correct sum through phase symmetry and frequency diversification.
Improves TTS accuracy by correcting context-dependent units.
problem Improves text-to-speech accuracy through speaker adaptation.
method Statistical model predicting context-dependent phonetic unit classes and their mean error values.
result Corrected boundaries of units improve TTS accuracy compared to HMM segmentation.
New method learns transport cost from subset correspondence.
problem Aligning multiple datasets with accurate cost functions.
method End-to-end optimizer (OT-SI) that differentiates through Sinkhorn algorithm.
result Substantially outperforms state-of-the-art benchmarks.
Enhances knowledge graph completion with mixed geometry tensor factorization.
problem Capturing nuanced distributional properties in knowledge graphs.
method Combines Euclidean and hyperbolic geometries for tensor factorization.
result Improves link prediction accuracy with fewer parameters.
Gradient descent trains both layers of a ReLU network to fit a linear model.
problem Training dynamics of a ReLU network to fit a linear target function.
method Jointly training both layers of a one-hidden-layer ReLU network in a realizable setting with Gaussian inputs and labels.
result Gradient descent from a small random initialization converges to a global minimizer at a linear rate with optimal sample complexity.
Researchers resolve a SYZ conjecture for A_n singularities using quantum-corrected T-duality.
problem Resolving a mathematically precise SYZ conjecture for A_n singularities.
method Building a quantum-corrected T-duality between two singular torus fibrations.
result Constructing a parameter-dependent SYZ mirror fibration partner with matching singular loci and integral affine structure.
In this paper, we describe a novel approach to imitation learning that infers latent policies directly from state observations. We introduce a method that characterizes the causal effects of latent actions on observations while simultaneously predicting their likelihood. We then outline an action alignment procedure th…
Method estimates LLM error rates using Pareto optimization.
problem Quantifying error rates in text-generating models.
method Pareto optimization for generating risk scores.
result Risk scores correlate well with true error rates.
New model simplifies symmetry handling in generative AI.
problem Symmetry handling in generative models for scientific tasks.
method Quotient-space diffusion models, viewing symmetry as quotient space.
result Improves performance over existing methods for molecular structure generation.
New covariance estimator for financial portfolios.
problem Estimating large financial covariances in non-stationary environments.
method Exponentially weighted averages and cross-validation for nonlinearly shrinking sample eigenvalues.
result Our estimator performs well in large dimensions compared to existing estimators.
Differentiable losses for combinatorial optimization problems in sequence modeling.
problem Mismatch between training and inference objectives in sequence models.
method Gradient descent over linear programs representing combinatorial optimization problems.
result Gradient descent can be applied to combinatorial optimization problems efficiently.
Improves classifier evaluation by aligning with Total Classification Cost.
problem Lack of consensus on evaluation metrics and class imbalance issues.
method Introduces Weighted Accuracy (WA) and a reweighting framework for cost-sensitive scenarios.
result WA aligns with Total Classification Cost (TCC) minimization under realistic conditions.
This new research explores the effects of various training methods on a Polish to English Statistical Machine Translation system for medical texts. Various elements of the EMEA parallel text corpora from the OPUS project were used as the basis for training of phrase tables and language models and for development, tunin…
FlashIV solves Black-Scholes implied volatility efficiently and accurately.
problem Efficiently calculating implied volatility for financial models.
method Normalizes inputs, uses asymptotic seed, and combines Householder refinement.
result Runs faster than existing methods while maintaining accuracy.
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.
The paper tackles temporal coverage bias in financial panel data, proposing a structuring framework to correct for incomplete histories.
problem Incomplete histories of financial instruments lead to biased panel data.
method Formalizes the problem and proposes a coverage-aware structuring framework using structured metadata and an availability matrix.
result The framework reveals substantial distortions in return dynamics and volatility when naive temporal alignment is used.