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.
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.
The paper proposes a method to align AI models using conformal risk control.
problem Aligning AI models to meet end-user requirements in non-generative settings.
method Post-processing a pre-trained model to better align with a subset of functions using conformal risk control.
result A probabilistic guarantee that the resulting conformal interval around a model contains a function approximately satisfying a desired property.
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.
Paper investigates monotonicity issues in AI preference learning.
problem AI models may violate monotonicity when learning preferences.
method Investigates root causes of non-monotonicity in comparison-based preference learning.
result Proves local pairwise monotonicity under mild assumptions.
PiNets provide faithful explanations for neural networks.
problem Lack of true explanations for neural network predictions.
method Pointwise-interpretable Networks (PiNets) that form linear models instance-wise.
result PiNets offer explanations that are meaningful, aligned, robust, and sufficient.
AI models aligned with human vision perform well on few data tasks.
problem Few-shot learning performance with limited data.
method Information-theoretic analysis and empirical testing of 491 models.
result Highly aligned models show better robustness to attacks and domain shifts.
Framework enhances AI explainability by aligning with human cognitive models.
problem Lack of explainability in AI models hinders trust and accountability.
method Integrates explainability techniques with Malle's five category model of behavior explanation.
result Demonstrates practical relevance in credit risk assessment and regulatory analysis.
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.
This paper introduces f-DPO, a generalized approach to Direct Preference Optimization using diverse divergence constraints.
problem Aligning large language models with human preferences while mitigating safety risks.
method Incorporates diverse divergence constraints to simplify the relationship between reward and optimal policy, eliminating the need for estimating the normalizing constant.
result Optimizes LLMs to align with human preferences more efficiently and under a broader set of divergence constraints.
Geometric framework detects concept frustration between human concepts and machine representations.
problem Aligning human concepts with machine learning representations.
method Geometric framework and similarity measures for detecting concept frustration.
result Concept frustration affects machine learning model performance and reorganizes learned concept representations.
This paper tests LLMs in finance to assess ethical behavior.
problem Aligning AI with ethical and legal standards in finance.
method Prompted LLMs to simulate CEO behavior, analyzed with logistic regression.
result Significant heterogeneity in LLMs' unethical behavior propensity.
QuantBench benchmarks AI methods for quantitative investment.
problem Lack of a standardized benchmark for AI in quantitative investment.
method Developed an industrial-grade benchmark platform with standardization, flexibility, and full-pipeline coverage.
result Revealed critical research directions in AI for quantitative investment.
Paper connects Plackett-Luce and Cox models for preference estimation.
problem Estimating preferences from annotated data.
method Connects Plackett-Luce model to Cox Proportional Hazards model.
result Implications of the connection between the two models.
AI can learn true probabilities if data and assumptions align.
problem Understanding when AI models can accurately represent true objective probabilities.
method Proved conditions under which AI can learn true probabilities.
result Conditions for learning true probabilities are identified.
Paper tackles hidden game problem in AI alignment and language games.
problem Hidden game problem in AI alignment and language games.
method Developed a composition of regret minimization techniques to discover and exploit hidden structures.
result Achieved optimal external and swap regret bounds for rapid convergence to correlated equilibria.
This work frames reward modelling from preferences as a causal problem.
problem Reward modelling from preference data for AI alignment.
method Causal inference approach to identify challenges and assumptions.
result Causally-inspired approaches improve model robustness.
Two impossibility theorems show formal alignment certification is impossible for AI systems.
problem Formal certification of AI alignment over open-ended domains is impossible.
method Two independent impossibility theorems: Semantic and Statistical barriers.
result No procedure can simultaneously satisfy soundness, completeness, and tractability.
Paper detects biases in medical imaging ML models using counterfactual analysis.
problem Bias in medical imaging ML models negatively impacts generalization performance.
method Counterfactual invariance framework combining conditional latent diffusion models and statistical hypothesis testing.
result The method identifies and quantifies biases without direct access to counterfactual data.
OOM-RL uses financial market losses to align AI agents in autonomous systems.
problem Constrained alignment of autonomous software agents in live financial markets.
method Deploying agents in live financial markets to enforce strict test-driven workflows.
result Final OOM-RL-aligned system achieved a stable equilibrium with an annualized Sharpe ratio of 2.06.
The paper analyzes risk spillovers between AI ETFs, AI tokens, and green markets.
problem Risk spillovers among AI ETFs, AI tokens, and green markets.
method R2 decomposition method
result AI ETFs and clean energy act as risk transmitters, while AI tokens and green assets act as receivers.
Paper addresses reward hacking in preference optimization, proposing POWER-DL to improve AI alignment.
problem Reward hacking problem in preference optimization, leading to undesired behaviors.
method POWER-DL combines robust reward maximization and dynamic label updates to mitigate reward hacking.
result POWER-DL consistently outperforms state-of-the-art methods on alignment benchmarks.
GAICF proposes a framework for governing generative AI in banking.
problem Generative AI's impact on financial decision-making and governance.
method SR 26-2-compatible governance framework for generative AI applications.
result GAICF aligns generative AI practices with SR 26-2 supervisory expectations.
GAICF proposes a framework for managing generative AI risks in banking.
problem Generative AI's impact on financial decision-making and governance.
method SR 26-2-compatible governance framework for generative AI.
result GAICF aligns generative AI practices with SR 26-2 supervisory expectations.
Sparse text alignments learned via optimal transport improve model explainability.
problem Building self-explaining models by selecting relevant text pieces.
method Employing optimal transport to find minimal cost alignments, introducing constrained variants for sparsity.
result Sparse and interpretable alignments achieved, preserving prediction accuracy.
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.
FST.ai 2.0 improves Taekwondo decision-making with AI, reducing review time and increasing trust.
problem Fair, transparent, and explainable decision-making in Taekwondo.
method Pose-based action recognition, epistemic uncertainty modeling, interactive dashboards.
result 85% reduction in decision review time, 93% referee trust in AI-assisted decisions.
The paper addresses the difficulty of decision makers trusting AI-assisted predictions and proposes a method to improve confidence values.
problem Decision makers struggle to trust AI-assisted predictions based on confidence values.
method The paper investigates why decision makers have difficulties and proposes a method to construct more useful confidence values.
result Multicalibration with respect to the decision maker's confidence on her own predictions is a sufficient condition for alignment, leading to better decisions.
Enhances AI models with human feedback for noisy data.
problem Improving AI model alignment with human feedback in noisy environments.
method Two-stage SL+LHF framework connecting machine learning with human feedback.
result The LNCA ratio identifies conditions for SL+LHF superiority over pure SL.
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.
RLHF uses human feedback to train AI models, posing statistical challenges.
problem Aligning AI models with human preferences using noisy, subjective feedback.
method Supervised fine-tuning, reward modeling, policy optimization, statistical ideas.
result Statistical methods for reward function learning and policy optimization.
New framework replicates private equity performance using AI and liquid strategies.
problem Inadequate trust and transparency in private equity markets.
method Advanced graphical models and asymmetric risk adjustments.
result Liquid, scalable solution that closely mimics private equity performance.
New AI models improve financial hedging by reducing shortfall and tail risk.
problem Static model calibration gaps in derivatives markets.
method Two reinforcement learning frameworks: RLOP and QLBS.
result RLOP reduces shortfall frequency and improves tail risk in stress scenarios.
Research tackles distribution shift issues in ML to improve AI reliability.
problem Distribution shift limits ML reliability and trustworthiness.
method Study three distribution shifts (perturbation, domain, modality) and investigate robustness, explainability, adaptability.
result Proposes effective solutions and fundamental insights for enhancing ML robustness, adaptability, and safety.
AutoML enhances credit decisions with XAI for better transparency.
problem Transparency in AI-driven financial decisions.
method Combining AutoML and XAI (SHAP) for credit scoring.
result Improved efficiency and accuracy in credit decisions with enhanced transparency.
Randomly initialized wide neural networks with zero-mean activations are nearly independent, potentially solving AI interpretability limits.
problem Measuring the limits of AI interpretability.
method Randomly initialized neural networks with large width and zero-mean activation functions.
result Neural networks with zero-mean activations are nearly independent, solving the computational no-coincidence conjecture.
FedSight AI predicts federal funds rate using LLMs and multi-agent reasoning.
problem Predicting Federal Open Market Committee's decisions on federal funds rate.
method Multi-agent framework with large language models, structured and unstructured inputs, and CoD extension for efficient reasoning.
result Achieved 93.75% accuracy and 93.33% stability in predicting FOMC outcomes.
This thesis tackles bias in AI decision-making in banking.
problem Bias in AI-driven banking decisions.
method Understanding, mitigating, and accounting for bias in AI systems.
result Establishment of Responsible AI practices for fair decision-making.
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 λ(f) 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.
Formalizes vNM utility theorem using Lean 4, proving existence and uniqueness.
problem Formalizing and proving the von Neumann-Morgenstern utility theorem.
method Implement classical axioms in Lean 4, formalizing preference relations over lotteries.
result Machine-verified proofs of existence and uniqueness of utility representations.
SEMASIA provides a large dataset of latent representations for model comparison.
problem Difficulty in comparing semantic structures across different neural network models.
method Collection of latent representations from 1700 pretrained models across various benchmarks.
result Consistent semantic organization across models and datasets.
Research creates a taxonomy to bridge AI security and regulatory gaps.
problem Disciplinary disconnect between technical and legal teams in AI risk assessment.
method Developed an AI System Threat Vector Taxonomy with 9 domains and 53 sub-threats.
result Empirically validated and aligned with ISO/IEC 42001 controls and NIST AI RMF functions.
FreB protocol uses AI to infer hidden parameters with valid confidence regions.
problem Generating biased or overconfident conclusions from AI-generated posterior distributions.
method Frequentist-Bayes (FreB) protocol reshapes AI-generated posterior distributions into valid confidence regions.
result FreB provides valid confidence regions that consistently include true parameters with expected probability.
ProEval efficiently estimates AI performance and discovers failures using pre-trained Gaussian Processes.
problem Resource-intensive evaluation of generative AI models.
method ProEval uses pre-trained Gaussian Processes and Bayesian quadrature to estimate performance and discover failures.
result ProEval requires significantly fewer samples to achieve accurate performance estimates and reveals more diverse failure cases.
RESTA defends LLMs against jailbreaking attacks by adding random noise to embeddings.
problem Vulnerability of LLMs to jailbreaking attacks that generate harmful outputs.
method Adds random noise to embedding vectors and aggregates during token generation.
result RESTA achieves superior robustness versus utility tradeoffs compared to baseline defenses.
Paper reviews the evolution of alpha from human insight to AI-powered systems.
problem Exceeding market benchmarks in finance.
method Five-stage taxonomy integrating representation learning, multimodal data fusion, and LLM agents.
result Unified framework for evaluating and developing next-gen alpha systems.
1-Lipschitz neural networks produce clearer, more focused Saliency Maps for explainable AI.
problem Noisy and limited Saliency Maps from traditional neural networks.
method Dual loss of optimal transport problem for 1-Lipschitz neural networks.
result Saliency Maps from 1-Lipschitz networks are highly concentrated and less noisy, aligning with human explanations.
New method detects bias in AI models that generate data.
problem Detecting bias in AI models that generate data.
method Formalized causal fairness in generative AI, derived new decomposition results, established identification conditions, and introduced efficient estimators.
result Demonstrated the value of new methodology in analyzing bias in large language models.