Transformers learn multi-step reasoning through gradient descent.
problem Understanding how transformers solve symbolic multi-step reasoning tasks.
method Theoretical analysis of gradient descent dynamics and multi-phase training.
result Trained one-layer transformers can solve both backward and forward reasoning tasks with generalization guarantees.
This work analyzes CoT prompting methods from a statistical estimation perspective.
problem Improving the effectiveness of LLMs in solving multi-step reasoning problems.
method Introducing a multi-step latent variable model to characterize CoT prompting from a statistical estimation viewpoint.
result The CoT estimator is equivalent to a Bayesian estimator when the pretraining dataset is large.
Achieving artificial visual reasoning - the ability to answer image-related questions which require a multi-step, high-level process - is an important step towards artificial general intelligence. This multi-modal task requires learning a question-dependent, structured reasoning process over images from language. Stand…
Looped Transformers learn to implement multi-step gradient descent for in-context learning.
problem Understanding the learnability of multi-step algorithms in multi-layer Transformers.
method Training weight-sharing looped Transformers for in-context linear regression, proving gradient dominance condition for convergence.
result Looped Transformers implement multi-step preconditioned gradient descent, converging to global minimizer.
MAC Net is a compositional attention network designed for Visual Question Answering. We propose a modified MAC net architecture for Natural Language Question Answering. Question Answering typically requires Language Understanding and multi-step Reasoning. MAC net's unique architecture - the separation between memory an…
We introduce a general-purpose conditioning method for neural networks called FiLM: Feature-wise Linear Modulation. FiLM layers influence neural network computation via a simple, feature-wise affine transformation based on conditioning information. We show that FiLM layers are highly effective for visual reasoning - an…
Most research in reading comprehension has focused on answering questions based on individual documents or even single paragraphs. We introduce a neural model which integrates and reasons relying on information spread within documents and across multiple documents. We frame it as an inference problem on a graph. Mentio…
Hydra boosts efficiency for long-context reasoning in resource-constrained settings.
problem Quadratic complexity of transformers limits long-context reasoning in resource-constrained systems.
method Hydra uses a modular architecture with adaptive routing between sparse global attention, mixture-of-experts, and dual memories.
result Hydra achieves significant throughput and accuracy improvements for long-context reasoning.
Study reveals how depth of reasoning affects generalization in models.
problem Understanding scaling behavior of generalization with CoT depth.
method Theoretical model of CoT in linear regression using random matrix theory.
result Sharp phase transition between exponential and polynomial improvement, saturation, and overthinking.
Transformers solve parity problems efficiently with step-by-step reasoning.
problem Training transformers to solve complex, recursive problems like parity.
method Training a one-layer transformer to solve k-parity, incorporating intermediate parities into the loss function, and using teacher forcing or augmented data. result Transformers can learn parity in one gradient update with intermediate supervision or self-consistency checks.
ChatGPT improves financial reasoning, overcoming biases in gold investment.
problem Improving financial reasoning and overcoming biases in investment decisions.
method Applied advanced prompt engineering and semantic news information to enhance LLMs' performance.
result ChatGPT with CoT prompt provides more explainable predictions and higher investment returns.
The paper calculates prices for multi-step barrier options under the Black-Scholes model.
problem Calculating prices for multi-step barrier options with varying barriers and time steps.
method Derives a general, explicit expression for option prices using the Black-Scholes model and a multi-step reflection principle.
result Derives a multi-step reflection principle that generalizes the reflection principle of Brownian motion.
Efficiently optimizes expensive functions with multi-step lookahead using one-shot optimization.
problem Optimizing expensive functions with long-term impacts using myopic approaches.
method Formulated as nested optimization problems within a multi-step scenario tree, optimized in one-shot fashion.
result Multi-step expected improvement is computationally tractable and outperforms existing methods.
Paper presents a copula-based method to efficiently generate correlated sample paths from multi-step time series models.
problem Generating realistic correlation structures in multi-step forecast sample paths is expensive and time-consuming.
method Copula-based approach to generate correlated sample paths in one forward pass.
result Improved sample path quality and significant speedup over autoregressive sampling.
The paper explores dynamic ensembles for multi-step forecasting.
problem Lack of research on dynamic ensembles for multi-step forecasting.
method Extensive experiments with 3568 time series and an ensemble of 30 multi-output models.
result Dynamic ensembles based on arbitrating and windowing perform best.
Reinforcement learning has attracted great attention recently, especially policy gradient algorithms, which have been demonstrated on challenging decision making and control tasks. In this paper, we propose an active multi-step TD algorithm with adaptive stepsizes to learn actor and critic. Specifically, our model cons…
Paper adapts ACI for online multi-step time-series forecasting with coverage guarantees.
problem Achieving reliable error bounds in online multi-step time-series forecasting.
method Adaptive conformal inference (ACI) adapted for multi-step forecasting with dynamic significance levels.
result Proposes a multi-step ACI algorithm with finite-sample coverage guarantees for non-exchangeable data.
Active-GRPO improves molecular optimization by actively deciding when to imitate or self-improve.
problem Training robust and efficient molecular optimization models with large language models.
method Active-GRPO combines imitation and reinforcement learning, upgrading references and policies dynamically.
result Improves molecular optimization performance, achieving statistically significant gains.
FinMaster benchmarks LLMs in financial tasks, revealing gaps in reasoning.
problem Challenges in financial tasks, including labor-intensive processes and low error tolerance.
method Developed a comprehensive financial benchmark (FinMaster) with three modules: FinSim, FinSuite, and FinEval.
result LLMs struggle with complex financial reasoning, showing significant accuracy drops.
Stratify unifies and improves multi-step forecasting strategies.
problem Lack of unified frameworks for multi-step forecasting strategies.
method Proposes Stratify, a parameterized framework for multi-step forecasting.
result Novel strategies in Stratify outperform existing ones in over 84% of experiments.
Multi-step passenger demand forecasting is a crucial task in on-demand vehicle sharing services. However, predicting passenger demand over multiple time horizons is generally challenging due to the nonlinear and dynamic spatial-temporal dependencies. In this work, we propose to model multi-step citywide passenger deman…
When environmental interaction is expensive, model-based reinforcement learning offers a solution by planning ahead and avoiding costly mistakes. Model-based agents typically learn a single-step transition model. In this paper, we propose a multi-step model that predicts the outcome of an action sequence with variable …
The paper introduces a multi-step loss function to improve model-based reinforcement learning.
problem Compounding of one-step prediction errors in long trajectories.
method A multi-step objective function combining MSE losses at various future horizons.
result Models trained with the multi-step loss achieve significant improvement in future prediction.
Knowledge graph reasoning is a critical task in natural language processing. The task becomes more challenging on temporal knowledge graphs, where each fact is associated with a timestamp. Most existing methods focus on reasoning at past timestamps and they are not able to predict facts happening in the future. This pa…
Multi-step temporal difference (TD) learning is an important approach in reinforcement learning, as it unifies one-step TD learning with Monte Carlo methods in a way where intermediate algorithms can outperform either extreme. They address a bias-variance trade off between reliance on current estimates, which could be …
Paper proposes a dual-level approach for multi-step forecasting of dynamical systems.
problem Accurate multi-step forecasting of time series systems for automatic control and optimization.
method Hybrid input forecasting using LSTM-STMs and physics-informed neural networks (PINNs).
result Hybrid models achieve higher log-likelihood and lower MSE compared to conventional methods.
For the efficient compensation of fiber nonlinearity, one of the guiding principles appears to be: fewer steps are better and more efficient. We challenge this assumption and show that carefully designed multi-step approaches can lead to better performance-complexity trade-offs than their few-step counterparts.
New framework guarantees convergence of multi-step MAML.
problem Convergence of multi-step MAML in nonconvex settings.
method Developed a theoretical framework for two types of MAML objective functions.
result Guaranteed convergence rate and computational complexity for multi-step MAML.
Model-based reinforcement learning is an appealing framework for creating agents that learn, plan, and act in sequential environments. Model-based algorithms typically involve learning a transition model that takes a state and an action and outputs the next state---a one-step model. This model can be composed with itse…
New method optimizes costly functions with unknown costs and budget constraints.
problem Optimizing functions with unknown and heterogeneous evaluation costs under a budget constraint.
method Budgeted multi-step expected improvement acquisition function.
result Our method outperforms existing approaches in various synthetic and real problems.
Multi-step ahead forecasting is still an open challenge in time series forecasting. Several approaches that deal with this complex problem have been proposed in the literature but an extensive comparison on a large number of tasks is still missing. This paper aims to fill this gap by reviewing existing strategies for m…
This paper compares self-reflection and budget tuning for LLMs, revealing domain-specific performance gains.
problem Improving inference-time performance of LLMs without retraining, balancing quality, cost, and latency.
method Systematic comparison of self-reflection and budget tuning across mathematical reasoning and translation tasks, evaluating various LLMs and model families.
result Substantial domain-dependent variation in self-reflection effectiveness, with up to 220% performance gains in mathematical reasoning.
AI uses language models to find instrumental variables quickly.
problem Finding valid instrumental variables is a challenging and heuristic process.
method Uses large language models to search for new instrumental variables through narratives and counterfactual reasoning.
result Demonstrates the effectiveness of multi-step and role-playing prompting strategies for LLMs.
This paper improves learning complex functions with CoT supervision, reducing sample complexity.
problem Learning complex functions with multi-step reasoning.
method Develops a statistical theory linking CoT risk and end-to-end risk, using CoT information measure.
result CoT supervision can achieve significantly faster learning rates compared to standard E2E supervision.
In its simplest form, the traffic flow prediction problem is restricted to predicting a single time-step into the future. Multi-step traffic flow prediction extends this set-up to the case where predicting multiple time-steps into the future based on some finite history is of interest. This problem is significantly mor…
ForecastNet uses a time-variant deep feed-forward neural network for better multi-step-ahead time series forecasting.
problem Time-invariant architectures limit multi-step-ahead forecasting.
method ForecastNet employs a deep feed-forward architecture with time-variant parameters and interleaved outputs.
result ForecastNet outperforms other models on multi-step-ahead time series forecasting tasks.
Diffusion-VAE tackles multi-step stock price prediction with stochastic noise.
problem Challenges in multi-step stock price prediction due to stochasticity and target price sequence.
method Combines hierarchical VAE and diffusion probabilistic techniques for seq2seq stock prediction.
result D-Va model outperforms state-of-the-art solutions in prediction accuracy and variance.
A multi-step framework tackles online unsupervised domain adaptation with novel mean-target subspace computation.
problem Online unsupervised domain adaptation with unlabelled target data arriving sequentially.
method Multi-step framework with a novel mean-target subspace computation and temporal coherency consideration.
result Improved performance over previous approaches on four datasets.
Quantile deep learning improves time series prediction accuracy and uncertainty quantification.
problem Uncertainty in multi-step time series prediction.
method Developed a novel quantile regression deep learning framework for multi-step time series prediction.
result Integrating quantile loss function with deep learning provides additional predictions for selected quantiles without loss in accuracy.
Proposes a framework to quantify uncertainty in multi-step decision-making by LLMs.
problem Uncertainty quantification in multi-step decision-making scenarios of LLMs.
method A principled, information-theoretic framework decomposing uncertainty into internal and extrinsic components, and proposing UProp for efficient extrinsic uncertainty estimation.
result UProp significantly outperforms existing single-turn UQ baselines in multi-step decision-making benchmarks.
AEnbMIMOCQR generates robust multi-step ahead prediction intervals for time series data.
problem Generating reliable multi-step ahead prediction intervals for time series data.
method Adaptive ensemble batch multi-input multi-output conformalized quantile regression (AEnbMIMOCQR) based on conformal prediction principles.
result AEnbMIMOCQR provides close to exact coverage and robustness to distribution shifts.
Multi-step greedy policies have been extensively used in model-based reinforcement learning (RL), both when a model of the environment is available (e.g.,~in the game of Go) and when it is learned. In this paper, we explore their benefits in model-free RL, when employed using multi-step dynamic programming algorithms: …
Detects corruption in agentic models during execution.
problem Inconsistent context, retrieval errors, or adversarial inputs corrupt intermediate steps of reasoning chains.
method Analyzes token graphs induced by attention and computes spectral statistics to emit accept/reject signals.
result A single threshold on the high frequency energy ratio optimally detects context inconsistency in agentic models.
We investigate the multi-step prediction of the drivable space, represented by Occupancy Grid Maps (OGMs), for autonomous vehicles. Our motivation is that accurate multi-step prediction of the drivable space can efficiently improve path planning and navigation resulting in safe, comfortable and optimum paths in autonom…
Weight quantization is one of the most important techniques of Deep Neural Networks (DNNs) model compression method. A recent work using systematic framework of DNN weight quantization with the advanced optimization algorithm ADMM (Alternating Direction Methods of Multipliers) achieves one of state-of-art results in we…
Unified LLM-agent with RL improves financial trading performance.
problem LLMs struggle with complex, multi-step financial tasks.
method Fusion of LLMs and gradient-based RL for policy optimization.
result Improves LLM performance in trading and other financial tasks.
We present a numerical framework for approximating unknown governing equations using observation data and deep neural networks (DNN). In particular, we propose to use residual network (ResNet) as the basic building block for equation approximation. We demonstrate that the ResNet block can be considered as a one-step me…
JANET improves time series prediction with adaptive uncertainty regions.
problem Time series data's lack of exchangeability and multi-step prediction challenges.
method Proposes JANET, a framework for joint adaptive prediction regions with controlled error rates.
result Demonstrates superior performance in multi-step prediction tasks across diverse datasets.