Research
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

58117175233 · Jun 202019922001200920172026
48 results for Inference-Time Activation Energy

ELS framework improves safety alignment by dynamically steering LLMs towards helpful responses.

problem Over-Refusal in Aligned Large Language Models
method Fine-tuning free framework using an Energy-Based Model (EBM) to dynamically steer LLMs during inference.
result Extensive experiments show a significant reduction in false refusals (from 57.3% to 82.6%) while maintaining safety performance.

RNE provides a flexible framework for diffusion models, enabling inference-time control and energy-based training.

problem Insufficient knowledge of marginal densities in diffusion models.
method Introduces Radon-Nikodym Estimator (RNE) to reveal the connection between marginal densities and transition kernels.
result RNE delivers strong results in inference-time control and energy-based diffusion training.

New method trains quantized neural networks to global optimality.

problem Training optimal quantized neural networks is intractable due to combinatorial non-convex optimization.
method Convex optimization strategy using hidden convexity, semidefinite lifting, and Grothendieck's identity.
result Quantized NN problems can be solved to global optimality in polynomial-time.

Active inference minimizes expected free energy for optimal behavior.

problem Understanding and optimizing behavior in complex systems.
method Combines Bayesian decision theory, optimal Bayesian design, and the free energy principle.
result Active inference emerges as a unified framework for information-seeking, utility maximization, and goal-directed behavior.

DriftLite improves inference quality of diffusion models without retraining.

problem Adapting pre-trained diffusion models to new target distributions without retraining.
method Lightweight, training-free particle-based approach that steers inference dynamics with optimal stability control.
result Consistently reduces variance and improves sample quality over existing methods.

LAD-BNet improves real-time energy forecasting on edge devices.

problem Real-time energy forecasting on edge devices for smart grid optimization and intelligent buildings.
method Hybrid neural architecture combining temporal lag exploitation and TCN with dilated convolutions.
result 14.49% MAPE at 1-hour horizon with 18ms inference time on Edge TPU, 8-12x faster than CPU.

Residential homes constitute roughly one-fourth of the total energy usage worldwide. Providing appliance-level energy breakdown has been shown to induce positive behavioral changes that can reduce energy consumption by 15%. Existing approaches for energy breakdown either require hardware installation in every target ho…

2019-09-02abs ↗pdf ↗

Bitcoin mining costs remain stable despite increased activity.

problem Stability of Bitcoin mining costs despite rising energy consumption.
method Estimation of global energy cost over a decade, considering changing oil costs and technological improvements.
result The cost of Bitcoin mining has remained relatively stable since 2010, relative to the volume of transactions.

Enhances drug discovery models by understanding human language.

problem Low predictive quality of activity prediction models in drug discovery.
method Proposes a novel architecture with separate chemical and natural language input modules and a contrastive pre-training objective.
result Improves predictive performance on few-shot and zero-shot learning benchmarks.

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.

Deep learning models accurately recognize and estimate physical activity types and energy expenditure from wrist accelerometer data.

problem Rigorous evaluation of wrist-worn accelerometers for assessing physical activity across the lifespan.
method Built deep learning networks to extract spatial and temporal representations from time-series data, recognizing physical activity types and estimating energy expenditure.
result Deep learning models achieved high performance: F1 scores of 0.82, 0.81, and 95 for sedentary, locomotor, and lifestyle activities, respectively; root mean square error of 1.1 for EE estimation.

We show LLMs can be locally linear, enabling better control of activations.

problem Suboptimal control of LLM activations during generation.
method Model LLM inference as a linear dynamical system, compute feedback controllers using Jacobians, and adapt classical control theory.
result Robust, fine-grained control of LLM activations across models and tasks.

Unified framework for planning under uncertainty using variational inference.

problem Planning under uncertainty with separate objectives for exploration and exploitation.
method Variational inference on a generative model augmented with priors.
result EFE-based planning emerges as variational inference, enabling scalable, resource-aware policies.

Tutorial on optimizing diffusion model samples for specific metrics.

problem Optimizing diffusion model samples for specific downstream metrics.
method Review and exploration of inference-time guidance and alignment methods.
result Unified perspective on inference-time algorithms and novel methods.

Active inference enhances RL by balancing exploration and exploitation.

problem Traditional RL's balance between exploration and exploitation is often suboptimal.
method Developed a new decision-making objective based on active inference.
result The new algorithm successfully balances exploration and exploitation on various RL benchmarks.

This paper introduces a new approach to active inference using constrained Bethe Free Energy.

problem Tackling the limitations of existing epistemic behavior models in active inference.
method Introducing a constrained Bethe Free Energy (CBFE) perspective to optimize epistemic behavior in generative models.
result CBFE optimization leads to more robust and flexible epistemic behavior compared to existing methods.

BALLAST optimizes Lagrangian observer placement for ocean vector fields.

problem Optimizing Lagrangian observer placement for time-dependent ocean vector fields.
method Bayesian active learning with look-ahead amendment for sea-drifter trajectories using a physics-informed spatio-temporal Gaussian process surrogate model.
result Noticeable benefits of BALLAST-aided observer placement strategies on synthetic and high-fidelity ocean models.

A DRL-based strategy improves vehicle tracking accuracy while saving energy.

problem Enhancing vehicle tracking accuracy in WSNs without increasing energy consumption.
method Decentralized strategy with dynamic reinforcement learning to adjust sensing areas.
result Simulation results demonstrate superior performance of DRL-aided design.

Large Language Models (LLMs) have demonstrated impressive generalization capabilities across various tasks, but their claim to practical relevance is still mired by concerns on their reliability. Recent works have proposed examining the activations produced by an LLM at inference time to assess whether its answer to a …

2025-06-10abs ↗pdf ↗

Unified Bayesian model explains in-context learning and activation steering in LLMs.

problem Understanding and controlling the behavior of large language models (LLMs) through prompts and activations.
method Developed a Bayesian model to explain and predict the effects of in-context learning and activation steering.
result Unified model predicts distinct phases and sudden shifts in LLM behavior, explaining prior empirical phenomena.

Generative model predicts daily activity sequences with duration-aware dynamics.

problem Accurately forecasting granular daily activity sequences for energy demand.
method Hierarchical semi-Markov models with duration-aware dynamics.
result Explicitly modeling activity durations improves predictive performance.

This work shows how approximate reward models can significantly improve inference-time scaling.

problem Improving the efficiency of inference for large language models.
method Identifying the Bellman error of approximate reward models and using Sequential Monte Carlo (SMC) for inference.
result Approximate reward models can reduce computational complexity from exponential to polynomial in TT.

New algorithm improves inference-time alignment without reward hacking.

problem Improving quality of responses from language models with limited compute.
method Inference-time alignment, focusing on extttInferenceTimePessimism exttt{InferenceTimePessimism} algorithm.
result Optimal performance and scaling-monotonicity of extttInferenceTimePessimism exttt{InferenceTimePessimism}.

EHR-MPC optimizes sepsis treatment using digital twins and inference-time control.

problem Optimal sepsis treatment policies are contested and difficult to adapt during inference.
method EHR-MPC decouples learning patient dynamics from treatment optimization, enabling inference-time control over learned digital twins.
result EHR-MPC achieves comparable off-policy performance and improved simulation performance compared to RL baselines.

Poisson variational autoencoders introduce a metabolic cost term that penalizes high baseline activity.

problem Energy constraints in computation.
method Poisson variational autoencoders with a Kullback-Leibler divergence term proportional to firing rates.
result Poisson variational autoencoders introduce a metabolic cost term that penalizes high baseline activity.

The problem of keyword spotting i.e. identifying keywords in a real-time audio stream is mainly solved by applying a neural network over successive sliding windows. Due to the difficulty of the task, baseline models are usually large, resulting in a high computational cost and energy consumption level. We propose a new…

2018-11-16abs ↗pdf ↗

New method combines deep learning and quantum mechanics for efficient molecular statistics.

problem Computational expense in extracting statistics from molecular systems.
method Adaptive Markov chain Monte Carlo with Normalizing Flow and MLP for quantum accuracy.
result Rapid convergence to Boltzmann distribution and accurate thermodynamic observables.

Active inference is a process theory of the brain that states that all living organisms infer actions in order to minimize their (expected) free energy. However, current experiments are limited to predefined, often discrete, state spaces. In this paper we use recent advances in deep learning to learn the state space an…

2020-01-30abs ↗pdf ↗

FEPS models agents to learn and act in complex environments without deep neural networks.

problem Modeling complex adaptive systems and understanding self-organizing behavior.
method Introducing Free Energy Projective Simulation (FEPS) within the constraints of the free energy principle and active inference.
result FEPS agents resolve ambiguity and infer optimal policies in partially observable environments.

DTS improves inference-time alignment of diffusion models with less compute.

problem Inference-time alignment of diffusion models suffers from inaccurate value estimation and inefficient reuse of past computations.
method Diffusion Tree Sampling (DTS) uses a tree-based approach to propagate terminal rewards and iteratively refine value estimates.
result DTS produces asymptotically exact samples and matches the FID of best-performing baselines with up to 10x less compute.

New theory shows predictive coding makes learning landscape easier to navigate.

problem Understanding the impact of predictive coding's inference procedure on learning efficiency.
method Analyzed the geometry of the energy landscape of deep linear networks, proving many non-strict saddles become strict in the equilibrated energy.
result All highly degenerate (non-strict) saddles of the loss become strict in the equilibrated energy, suggesting a more robust learning landscape.

URGE improves diffusion model quality without gradients or Hessian.

problem Improving sample quality in diffusion models without gradient evaluations.
method Path-wise importance reweighting via Girsanov change of measure.
result URGE achieves better generation quality than existing methods.

Deep networks run with low precision operations at inference time offer power and space advantages over high precision alternatives, but need to overcome the challenge of maintaining high accuracy as precision decreases. Here, we present a method for training such networks, Learned Step Size Quantization, that achieves…

2019-02-21abs ↗pdf ↗

New method aligns diffusion models for inference-time properties without retraining.

problem Aligning pre-trained diffusion models for desired inference-time properties.
method Variationally stable Doob's matching for provable guidance estimation.
result Consistent estimator of guidance with non-asymptotic convergence guarantees.

Agents learn and control complex mechanical systems through shared memories.

problem Controlling multi-joint dynamical systems.
method Coupled autoregressive active inference agents using Bayesian filtering and minimizing expected free energy.
result Demonstrated learning and control of a double mass-spring-damper system.

Artificial neural networks typically have a fixed, non-linear activation function at each neuron. We have designed a novel form of piecewise linear activation function that is learned independently for each neuron using gradient descent. With this adaptive activation function, we are able to improve upon deep neural ne…

2014-12-21abs ↗pdf ↗