All papers

Introduces a benchmark to test model generalization under systematic convenience sampling biases in tabular data.

The paper introduces an agentic knowledge graph retrieval framework that uses vulnerability taxonomies to generate secure code.

The paper proposes an execution-gating method to safely launch and cancel tool calls early during live speech recognition before user self-corrections.

The paper presents a saliency-guided patch replay method to mitigate catastrophic forgetting in spiking neural networks under tight memory budgets.

Presents a training-free framework combining variance-adaptive particle search with Metropolis-Hastings refinement for budget-constrained audio-visual alignment.

The paper introduces QuasiLeak-Bench to evaluate whether privacy detectors can identify distributed quasi-identifiers governed by disclosure policies.

The paper presents techniques to adapt and quantize a small language model for on-device deployment in phone call agent services.

The paper introduces signed trajectories and runtime verification to prevent tampered workflows from corrupting shared agent workflow repositories.

The authors design an FPGA implementation of spiking neural networks that performs both inference and training on-chip without backpropagation through time.

Introduces a shift-aware identity-guided latent refinement framework using dual tokens to resolve target drift in referring audio-visual segmentation.

The paper introduces ExecutionRAG to show how search drift causes apparent retrieval failures in multi-hop question answering.

The paper introduces a dual-process memory framework coordinating small and large models for efficient long-context reasoning.

This paper introduces a realistic benchmark of 132 business-logic vulnerabilities to evaluate cybersecurity AI agents without contamination.

The paper develops a plug-and-play module for convolutional neural networks that provides optimal rotation and translation invariance based on invariant theory.

Generates seamlessly looping 3D cinemagraphs from static 3D Gaussian splats using periodic deformation fields guided by video models.

Compares abstract-only versus full-text inputs and few-shot versus fine-tuning approaches for generating biomedical lay summaries.

The paper evaluates the ability of language model agents to diagnose appliance faults using a simulated appliance benchmark based on sensor data.

This paper investigates membership inference attacks on vision-language models using low-noise Gaussian perturbations to detect training data exposure.

The paper proposes a rubric-based reinforcement learning method combining selective teacher distillation with an adaptive hand-off to standard optimization.

The paper proposes a complex-valued neural network to fuse audio and visual similarity streams for open-vocabulary audio-visual event localization.

Shows across 200 models that proposed intrinsic tokenizer metrics do not reliably predict downstream language model performance.

Proposes a structured knowledge admission and verification system to prevent continual learning agents from storing harmful past experience entries.

This paper proposes a tree-guided reinforcement learning framework to discover and probe hierarchical jailbreak strategies against language models.

The paper applies Hilbert geometry to constrain activation shifts in softmax layers for stable optimizer updates in language model heads.

The paper proposes MIDAPN, a unified spatiotemporal network using identity-aware graphs and temporal learning across numerical and narrative-flow time series.

The paper introduces a benchmark to evaluate proactive video assistants across perception, risk prediction, and timely real-world intervention tasks.

The paper proposes a framework that replaces generative model calls with lightweight decision models for routing and reviewing in multi-agent workflows.

This paper analyzes vulnerabilities in system interfaces where adapter substitutions alter model decisions without modifying the underlying model itself.

Proposes Dynamic K-Head, a question-conditioned budget selector that sets continuous latent reasoning steps based on input difficulty.

Proposes a reinforcement learning framework with verifiable rewards and error grounding to align multimodal emotion descriptions with human preferences.

The paper introduces a benchmark to evaluate question answering systems on reasoning and predictive tasks over relational databases.

The paper presents a framework to evaluate whether language model agents can build lower-cost benchmark tests that match reference evaluations.

This paper demonstrates an electromagnetic side-channel attack that extracts generated tokens and user prompts from edge devices running language models.

The paper develops and evaluates a multimodal foundation model for interpreting oncologic chest, abdomen, and pelvis CT scans.

The paper introduces a direction- and topology-aware graph transformer to predict neural network architecture attributes like accuracy and latency.

The paper proposes a policy optimization method that treats reasoning transitions as learned coordinates in an energy space to internalize self-correction.

This paper studies prompt engineering tactics aimed at bypassing multi-layer agent security defenses across diverse enterprise tasks.

Introduces a calibration-free latent merging framework to compress Mixture-of-Experts models and reduce memory overhead.

The paper presents DVLA-RL++, combining complementary semantic purification and reinforcement learning gating to reduce context contamination in few-shot learning.

Introduces AAArena, a benchmark evaluating how AI agents revise executable code policies across 12 adversarial games under fixed budgets.

The paper introduces a hierarchical benchmark for evaluating visual generative models on multi-step world state prediction tasks.

Introduces a self-supervised consistency loss across varied tool interfaces to make language model agents robust against notation and naming changes.

This paper uses loss curvature to identify and remove fragile directions that cause activation steering vectors to fail.

Proposes a structured zeroth-order optimization framework using subspace estimation and spectral control for memory-efficient model fine-tuning.

The paper adapts a causal language model using compressed 3D geometric tokens to assemble multi-part objects in a single forward pass.

Proposes a visual verifier that decomposes complex prompts into sub-questions and assigns continuous scores to improve multimodal evaluation.

The paper proposes a runtime monitoring method to detect and correct language model agents that deviate from intended procedures without reporting it.

This paper presents an unsupervised test-time adaptation method that uses movement-induced prediction variance to adapt object recognition models in mobile agents.

This paper presents a method to detect spoken code-switching directly from speech representations using language-selective hidden dimensions.

Presents a loop-aware post-training quantization method to compress looped language models without severe accuracy loss.

The paper measures whether requester users can understand multi-step agent replies and evaluates how explicit constraints improve reply usability.

This paper analyzes multi-person conversational involvement by jointly evaluating speaking floors and spatial formations from multimodal video features.

The paper evaluates a framework that replaces spoken patient identifiers and alters voice identity while preserving clinical conversational evidence.

Presents an activation quantization method that groups tokens via pre-scan features to improve low-bit post-training quantization for Mamba super-resolution models.

This paper introduces an agentic post-training framework that coordinates multiple agents to evaluate and adaptively train text-to-image models.

The paper learns trajectory embeddings to predict the success or failure of language model agents early during execution.

The paper proposes a distillation framework to remove specific multimodal knowledge without causing hallucinations or leaving residual queryable information.

The paper proposes a reasoning framework that discretizes visual states into tokens to jointly optimize visual computation and chain-of-thought steps.

The authors propose a resource-efficient two-stage approach for estimating 3D road user shapes and orientations from roadside cameras and radar.

The paper presents an importance-weighted downsampling protocol to lower evaluation costs on transaction graph datasets without distorting precision metrics.