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.
DTA aligns multimodal data with prior correspondence knowledge.
problem Challenges in aligning data from different domains.
method Semi-supervised manifold alignment method exploiting prior correspondence knowledge.
result DTA outperforms other methods in aligning multimodal data.
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.
Decision-alignment evaluates uncertainty quantification for decision-relevant UQ
problem Evaluation of uncertainty quantification metrics
method Introduce decision-alignment
result Proper scoring rules align with decision utility
New Gromov-Wasserstein metric controls rigidity and incorporates prior knowledge.
problem Inflexible Gromov-Wasserstein distance and lack of feature alignment.
method Augmented Gromov-Wasserstein distance with feature alignments and prior knowledge.
result Improved performance in single-cell multi-omic alignment and transfer learning.
We present a model that can automatically learn alignments between high-dimensional data in an unsupervised manner. Our proposed method casts alignment learning in a framework where both alignment and data are modelled simultaneously. Further, we automatically infer groupings of different types of sequences within the …
Bayesian HMM for protein alignment state estimation.
problem Estimating the MAP state sequence for a Bayesian HMM.
method Training data with Dirichlet priors, iterative algorithms for MAP path estimation.
result Bayesian setup outperforms frequentist estimation in protein alignment.
New meta-reinforcement learning method improves performance in finite-horizon MDPs.
problem Improving meta-reinforcement learning in finite-horizon MDPs with shared optimal action-value functions.
method Proposes MTSRL and MTSRL+ algorithms with learned priors and covariance, coupled with prior-alignment technique for meta-regret guarantees.
result Achieves meta-regret guarantees with learned priors and covariance, outperforming prior-independent RL and bandit-only meta-baselines.
Algorithm learns shared demand structure across dynamic pricing experiments.
problem Learning shared demand parameters across multiple dynamic pricing experiments.
method Meta dynamic pricing algorithm that learns prior online while solving Thompson sampling experiments.
result Algorithm achieves sublinear meta regret in experiment-rich environments.
New method for robust distribution alignment using log-likelihood ratio and normalizing flows.
problem Distribution alignment challenges in deep learning.
method Log-likelihood ratio statistic and normalizing flows.
result Minimizing the proposed objective yields robust domain alignment.
DRO-REBEL improves LLM alignment by robustly updating models online.
problem Overfitting and drifting of LLMs during RLHF.
method DRO-REBEL uses type-p Wasserstein, KL, and χ2 ambiguity sets for robust online updates. result DRO-REBEL achieves faster convergence and better performance than prior methods.
SAIL improves online alignment of large language models with minimal feedback.
problem Offline RLHF methods often lead to sub-optimal performance due to fixed preference datasets.
method SAIL uses bilevel optimization and a single-level first-order method to iteratively refine model alignment.
result SAIL significantly improves alignment performance on open-sourced datasets with minimal computational overhead.
Emergent misalignment is influenced by training dynamics, model priors, and data.
problem Emergent misalignment in models
method Exploring training dynamics, model priors, and data
result Activation deltas before and after narrow fine-tuning correlate with their similarities when measured with the last prompt-token activations.
Unified framework for imitating tasks across domains with discrepancies.
problem Learning tasks across domains with embodiment, viewpoint, and dynamics mismatches.
method Two-step approach: alignment followed by adaptation. Alignment uses Generative Adversarial MDP Alignment (GAMA) for state and action correspondences from unpaired, unaligned demonstrations. Adaptation leverages these correspondences for zero-shot imitation.
result Effectiveness of the proposed approach in embodiment, viewpoint, and dynamics mismatch scenarios.
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.
PAC-Bayesian theory improves text-to-image models by enforcing alignment and generalization.
problem Text-to-image models struggle with complex prompts, misaligning modifiers and neglecting certain elements.
method Proposes a Bayesian approach with custom priors over attention distributions to enforce desirable properties.
result Achieves state-of-the-art results across multiple metrics on standard benchmarks.
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.
Graph Neural Networks align with dynamic programming, improving algorithmic reasoning.
problem Demonstrate and quantify alignment between GNNs and dynamic programming.
method Category theory and abstract algebra methods to expose intricate connection.
result Showed GNNs align with dynamic programming beyond individual algorithms.
Proposes a non-adversarial method for distribution matching.
problem Stability and optimization challenges in adversarial matching methods.
method Non-adversarial VAE-based matching method with alignment upper bounds.
result Demonstrates applicability of non-adversarial matching methods without modifying original architectures.
This work explores how overparametrization and priors affect Bayesian neural network posteriors.
problem Symmetries, non-identifiabilities, and weight-space priors fragment and inflate BNN posteriors.
method We study the interplay between overparametrization and priors in BNN posteriors, deriving key phenomena and validating through experiments.
result Overparametrization induces structured, prior-aligned weight posterior distributions.
This paper argues for decolonizing AI alignment by incorporating open-source Hinduism concepts.
problem Coloniality in AI development and deployment, particularly in alignment practices.
method Proposes three forms of openness: model, societal, and excluded knowledge openness, using Hindu viśe\d{s}a-dharma.
result AI alignment should be decolonialized to avoid moral absolutism and better align with desired values.
A new machine learning method for Bayesian inverse problems in function spaces.
problem Bayesian inverse problems in function spaces with incompatibility of white noise sources.
method One-step generative transport with amortized neural operator and prior-aligned Gaussian random field.
result Generative operator trained on prior samples and noisy observations generates posterior samples efficiently.
This paper establishes risk convergence and asymptotic weight matrix alignment --- a form of implicit regularization --- of gradient flow and gradient descent when applied to deep linear networks on linearly separable data. In more detail, for gradient flow applied to strictly decreasing loss functions (with similar re…
UDA improves ABI robustness but fails under certain prior misspecifications.
problem Robustness of ABI in noisy real-world data.
method Systematic evaluation of UDA across various misspecification scenarios.
result UDA aligns summary spaces but can fail under prior misspecifications.
In this paper we introduce the deep kernelized autoencoder, a neural network model that allows an explicit approximation of (i) the mapping from an input space to an arbitrary, user-specified kernel space and (ii) the back-projection from such a kernel space to input space. The proposed method is based on traditional a…
The application of machine learning to bioinformatics problems is well established. Less well understood is the application of bioinformatics techniques to machine learning and, in particular, the representation of non-biological data as biosequences. The aim of this paper is to explore the effects of giving amino acid…
RAFT fine-tunes models using high-quality samples to align them with human preferences.
problem Aligning generative models with human ethics and preferences.
method RAFT selects high-quality samples, discards undesired behavior, and fine-tunes the model on filtered samples.
result RAFT improves model performance in reward learning and automated metrics.
Oblique BART improves tree-based predictions.
problem Axis-aligned decision rules in BART can be suboptimal.
method Developed an oblique version of BART using data-adaptive hyperplane partitions.
result Oblique BART outperformed axis-aligned BART and other tree methods on benchmarks.
New framework transfers latent knowledge from weak to strong models.
problem Aligning superhuman LLMs with human feedback.
method Transfer learning framework using refinement approach.
result Proves weak-to-strong generalization is possible.
Joint alignment of a collection of functions is the process of independently transforming the functions so that they appear more similar to each other. Typically, such unsupervised alignment algorithms fail when presented with complex data sets arising from multiple modalities or make restrictive assumptions about the …
CUDA CTDR tackles unsupervised domain adaptation without domain alignment.
problem Lack of direct methods for unlabeled target domain classification.
method Jointly learns CTDR on source and target distributions using contradistinguish loss and supervised loss.
result CUDA CTDR achieves state-of-the-art results on various domain adaptation datasets.
New method robustly discovers causal relationships from imperfect data.
problem Challenges in causal discovery from imperfect structural constraints.
method Prior alignment and conflict resolution through surrogate model and multi-task learning.
result Proposes a robust method for causal discovery under imperfect constraints.
TSFlow uses Gaussian processes to match priors for better time series forecasting.
problem Difficulties in aligning generative models' priors with time series data.
method Conditional flow matching (CFM) with Gaussian processes, optimal transport, and data-dependent priors.
result TSFlow produces high-quality unconditional samples and competitive forecasting results.
Deep Gaussian Processes with polynomial kernels can collapse rapidly without proper hyperparameter tuning.
problem The collapse of Deep Gaussian Processes with polynomial kernels without careful hyperparameter tuning.
method Analysis using the Berry-Esseen Theorem and observation of prior behavior.
result The prior of a Deep Gaussian Process collapses rapidly towards zero or places negligible mass on low norm functions without proper hyperparameter tuning.
Optimizes pruning masks for neural networks using probabilistic fine-tuning and PAC-Bayes bounds.
problem Improving neural network performance through adaptive pruning of weights.
method Optimizes stochastic pruning masks by minimizing expected loss, considering data-adaptive regularization and feature alignment.
result Probabilistic fine-tuning leads to improved test error over baseline methods in neural networks.
Bayesian neural networks incorporate domain knowledge through variational inference.
problem Specifying priors for Bayesian neural networks that capture domain knowledge is challenging.
method Proposes a framework for integrating domain knowledge into BNN priors through variational inference.
result BNNs with proposed domain knowledge priors outperform those with standard priors, achieving better predictive performance.
Improves text-to-image diffusion models using GFlowNets.
problem Aligning diffusion models with text descriptions.
method Post-training diffusion models with GFlowNets to generate high-reward images.
result Effective alignment of large-scale text-to-image diffusion models with reward information.
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.
Efficiently learns Single-Index Models with constant factor approximation.
problem Learning Single-Index Models under L22 loss with unknown link functions. method An efficient algorithm using alignment sharpness for optimization.
result Achieves constant factor approximation to optimal loss for various distributions and link functions.
GrokAlign aligns Jacobians to accelerate grokking in deep networks.
problem Accelerating the training dynamics of deep networks to avoid delayed generalisation and robustness.
method Aligning the Jacobians of a deep network with the training data to ensure grokking under a low-rank assumption.
result GrokAlign regularizes Jacobians to induce grokking sooner than conventional methods.
Gradient descent aligns neural feature matrices with pre-activation tangent features.
problem Understanding neural feature learning mechanisms.
method Analytical proof of alignment between weight matrices and pre-activation tangent features.
result Derivative alignment occurs almost surely in high-dimensional settings.
New method flattens decision boundary by targeting shortcut-aligned axes in disentangled latent space.
problem Shortcut learning in neural networks, leading to poor out-of-distribution generalization.
method Injects targeted anisotropic noise to regularize classifier sensitivity along shortcut-aligned axes.
result Achieves state-of-the-art OOD performance without shortcut labels or conflicting samples.
Method evaluates disentanglement in DLVMs, including those not aligned with latent axes.
problem Evaluate disentanglement in DLVMs, especially those not aligned with latent axes.
method Proposes a statistical method to discover generative factors of a dataset.
result Empirically demonstrates the advantage of the method on two datasets.
Adaptive optimal transport priors improve few-shot learning robustness.
problem Limited supervision and distribution shifts in few-shot learning.
method Prototype-Guided Distributionally Robust Optimization (PG-DRO) framework.
result PG-DRO achieves stronger robust generalization in few-shot scenarios.
ReMixMatch improves semi-supervised learning with new techniques for data efficiency.
problem Improving semi-supervised learning with limited labeled data.
method Distribution alignment and augmentation anchoring with AutoAugment.
result Significantly more data-efficient, requiring less labeled data for similar accuracy.
Paper proposes f-DPG for aligning language models with preferences.
problem Aligning language models with user preferences.
method Uses f-divergence to approximate target distributions and minimizes a forward KL from it using DPG.
result Jensen-Shannon divergence often outperforms forward KL divergence, leading to significant improvements.
Develops a simulation-based method to translate expert knowledge into prior distributions for Bayesian models.
problem Effective incorporation of expert knowledge into prior distributions for diverse model structures.
method Simulation-based stochastic gradient descent to learn hyperparameters of parametric priors from expert knowledge.
result Method is adaptable to various elicitation techniques and independent of model structure.
New method improves generative modeling on convex domains using regularized mirror maps and Student-t priors.
problem Challenges in generative modeling on convex domains with heavy-tailed targets.
method Mirror Flow Matching with regularized mirror maps and Student-t priors.
result Empirically outperforms baselines and achieves competitive sample quality.