X-Tokenizer creates semantic action tokens via asymmetric residual quantization and contrastive pretraining on large trajectory data, outperforming prior methods like FAST on robotic tasks.
hub Mixed citations
Bert: Pre-training of deep bidirectional transformers for language understanding
Mixed citation behavior. Most common role is background (60%).
29
Pith papers citing it
Background
60% of classified citations
hub tools
citation-role summary
background 7
method 2
baseline 1
citation-polarity summary
representative citing papers
SHADOWMASK backdoors MDLMs by replacing the all-mask terminal distribution with a trigger-mask mixture prior, achieving near-100% attack success on DiT and LLaDA-8B models across multiple datasets while resisting fine-tuning and some defenses.
MA-BC partitions divergent expert data and pools non-conflicting pairs to achieve faster convergence to Pareto-optimal policies in MOMDPs, with a matching minimax lower bound.
Optimal INR freeze depth matches highest weight stable rank layer; SAEs reveal SIREN atoms are localized while FFMLP atoms trace cohort contours with causal impact on PSNR.
AdaTKG equips temporal knowledge graph entities with per-entity memories updated via a single shared learnable exponential moving average, allowing online adaptation and better reasoning on evolving facts.
SHREC is a new benchmark dataset of embodied human-robot conversations that shows substantial performance gaps in state-of-the-art foundation models on tasks involving social error detection and rationale generation.
ER-JEPA, a lightweight two-stage Joint-Embedding Predictive Architecture that separates channel-wise and temporal processing, achieves state-of-the-art ECG classification on PTB-XL with lower compute than prior transformer models.
A tri-modal contrastive learning method for EEG-based zero-shot visual decoding reports 54.1% top-1 accuracy on the Things-EEG2 200-way benchmark, outperforming prior baselines of 32.4%.
ChunkFT enables full-parameter fine-tuning of Llama 3-8B on one 24 GB GPU and Llama 3-70B on two 80 GB GPUs by streaming gradients over dynamically activated sub-tensors.
HypergraphFormer trains LLMs via supervised fine-tuning to generate hypergraph textual representations for floor plans, claiming better performance than raster or vector methods on RPLAN and a new out-of-distribution dataset while enabling arbitrary boundaries and high editability.
A transformer-based in-context learning model predicts continental-scale subsurface temperatures from sparse borehole observations, outperforming physics and interpolation baselines while adapting to new regions with 20 examples.
Nonlinear Bipolar Compensation with Bipolar Logarithmic Transformation reduces outlier effects in post-training quantization by performing compensation in a compressed transformed space.
MILM fine-tunes LLMs on XML-encoded multimodal irregular time series via a two-stage process that exploits informative sampling patterns to achieve top performance on EHR classification datasets.
An audit of 152 papers reveals that geospatial foundation models lack standardized evaluations, training controls, and weight releases, so no one knows the state of the art.
RigidFormer learns mesh-free rigid dynamics from point clouds using object-centric anchors, Anchor-Vertex Pooling, Anchor-based RoPE, and differentiable Kabsch alignment to enforce rigidity.
A bilevel method learns composite pretraining loss weights online via gradient alignment with a downstream objective, matching tuned baselines at roughly 30% extra cost over one training run.
Intern-Atlas constructs a methodological evolution graph with 9.4 million edges from 1.03 million AI papers to capture how methods emerge, adapt, and transition, enabling better idea evaluation and generation for AI-driven research.
XAI explanations should be narratives with continuous structure, cause-effect, fluency and diversity, and new metrics are needed to evaluate this better than standard NLP scores.
The work introduces uncertainty-aware foundation models for clinical data by learning set-valued patient representations that enforce consistency across partial observations and integrate multimodal self-supervised objectives.
Causal2Vec prepends a BERT-generated contextual token to decoder-only LLMs and pools its hidden state with the EOS token to reach new SOTA on MTEB among public-data-trained embedding models.
CodeBrain introduces a decoupled TFDual-Tokenizer and multi-scale EEGSSM architecture for an EEG foundation model pretrained on a large corpus, claiming strong generalization across eight downstream tasks and ten datasets.
AID-VAR attaches an adversarial discriminator and lightweight guidance injector to frozen VAR backbones to diagnose and correct fidelity gaps across scales, reporting 16% FID gains with 3% added parameters.
SFT on LLMs removes noise-like token interactions in a brief early phase before introducing overfitted ones, explaining inconsistent effectiveness across model scales.
EHR-RAGp is a retrieval-augmented EHR foundation model that employs prototype-guided retrieval to dynamically integrate relevant historical patient context, outperforming prior models on clinical prediction tasks.
citing papers explorer
-
X-Tokenizer: A Multimodal Action Tokenizer for Vision-Language-Action Pretraining
X-Tokenizer creates semantic action tokens via asymmetric residual quantization and contrastive pretraining on large trajectory data, outperforming prior methods like FAST on robotic tasks.
-
Backdooring Masked Diffusion Language Models
SHADOWMASK backdoors MDLMs by replacing the all-mask terminal distribution with a trigger-mask mixture prior, achieving near-100% attack success on DiT and LLaDA-8B models across multiple datasets while resisting fine-tuning and some defenses.
-
Split the Differences, Pool the Rest: Provably Efficient Multi-Objective Imitation
MA-BC partitions divergent expert data and pools non-conflicting pairs to achieve faster convergence to Pareto-optimal policies in MOMDPs, with a matching minimax lower bound.
-
What Cohort INRs Encode and Where to Freeze Them
Optimal INR freeze depth matches highest weight stable rank layer; SAEs reveal SIREN atoms are localized while FFMLP atoms trace cohort contours with causal impact on PSNR.
-
AdaTKG: Adaptive Memory for Temporal Knowledge Graph Reasoning
AdaTKG equips temporal knowledge graph entities with per-entity memories updated via a single shared learnable exponential moving average, allowing online adaptation and better reasoning on evolving facts.
-
Social Human Robot Embodied Conversation (SHREC) Dataset: Benchmarking Foundational Models' Social Reasoning
SHREC is a new benchmark dataset of embodied human-robot conversations that shows substantial performance gaps in state-of-the-art foundation models on tasks involving social error detection and rationale generation.
-
Hierarchical Self-Supervised Representation Learning Framework for Multivariate Time Series Grounded in ECG Analysis
ER-JEPA, a lightweight two-stage Joint-Embedding Predictive Architecture that separates channel-wise and temporal processing, achieves state-of-the-art ECG classification on PTB-XL with lower compute than prior transformer models.
-
MindAlign: Bridging EEG, Vision, and Language for Zero-Shot Visual Decoding
A tri-modal contrastive learning method for EEG-based zero-shot visual decoding reports 54.1% top-1 accuracy on the Things-EEG2 200-way benchmark, outperforming prior baselines of 32.4%.
-
ChunkFT: Byte-Streamed Optimization for Memory-Efficient Full Fine-Tuning
ChunkFT enables full-parameter fine-tuning of Llama 3-8B on one 24 GB GPU and Llama 3-70B on two 80 GB GPUs by streaming gradients over dynamically activated sub-tensors.
-
HypergraphFormer: Learning Hypergraphs from LLMs for Editable Floor Plan Generation
HypergraphFormer trains LLMs via supervised fine-tuning to generate hypergraph textual representations for floor plans, claiming better performance than raster or vector methods on RPLAN and a new out-of-distribution dataset while enabling arbitrary boundaries and high editability.
-
In-context learning enables continental-scale subsurface temperature prediction from sparse local observations
A transformer-based in-context learning model predicts continental-scale subsurface temperatures from sparse borehole observations, outperforming physics and interpolation baselines while adapting to new regions with 20 examples.
-
Nonlinear Bipolar Compensation: Handling Outliers in Post-Training Quantization
Nonlinear Bipolar Compensation with Bipolar Logarithmic Transformation reduces outlier effects in post-training quantization by performing compensation in a compressed transformed space.
-
MILM: Large Language Models for Multimodal Irregular Time Series with Informative Sampling
MILM fine-tunes LLMs on XML-encoded multimodal irregular time series via a two-stage process that exploits informative sampling patterns to achieve top performance on EHR classification datasets.
-
No One Knows the State of the Art in Geospatial Foundation Models
An audit of 152 papers reveals that geospatial foundation models lack standardized evaluations, training controls, and weight releases, so no one knows the state of the art.
-
RigidFormer: Learning Rigid Dynamics using Transformers
RigidFormer learns mesh-free rigid dynamics from point clouds using object-centric anchors, Anchor-Vertex Pooling, Anchor-based RoPE, and differentiable Kabsch alignment to enforce rigidity.
-
When Losses Align: Gradient-Based Composite Loss Weighting for Efficient Pretraining
A bilevel method learns composite pretraining loss weights online via gradient alignment with a downstream objective, matching tuned baselines at roughly 30% extra cost over one training run.
-
Intern-Atlas: A Methodological Evolution Graph as Research Infrastructure for AI Scientists
Intern-Atlas constructs a methodological evolution graph with 9.4 million edges from 1.03 million AI papers to capture how methods emerge, adapt, and transition, enabling better idea evaluation and generation for AI-driven research.
-
On the Importance and Evaluation of Narrativity in Natural Language AI Explanations
XAI explanations should be narratives with continuous structure, cause-effect, fluency and diversity, and new metrics are needed to evaluate this better than standard NLP scores.
-
Uncertainty-Aware Foundation Models for Clinical Data
The work introduces uncertainty-aware foundation models for clinical data by learning set-valued patient representations that enforce consistency across partial observations and integrate multimodal self-supervised objectives.
-
Causal2Vec: Improving Decoder-only LLMs as Embedding Models through a Contextual Token
Causal2Vec prepends a BERT-generated contextual token to decoder-only LLMs and pools its hidden state with the EOS token to reach new SOTA on MTEB among public-data-trained embedding models.
-
CodeBrain: Bridging Decoupled Tokenizer and Multi-Scale Architecture for EEG Foundation Model
CodeBrain introduces a decoupled TFDual-Tokenizer and multi-scale EEGSSM architecture for an EEG foundation model pretrained on a large corpus, claiming strong generalization across eight downstream tasks and ten datasets.
-
Adversarial Error Correction for Visual Autoregressive Generation
AID-VAR attaches an adversarial discriminator and lightweight guidance injector to frozen VAR backbones to diagnose and correct fidelity gaps across scales, reporting 16% FID gains with 3% added parameters.
-
Reconciling Contradictory Views on the Effectiveness of SFT in LLMs: An Interaction Perspective
SFT on LLMs removes noise-like token interactions in a brief early phase before introducing overfitted ones, explaining inconsistent effectiveness across model scales.
-
EHR-RAGp: Retrieval-Augmented Prototype-Guided Foundation Model for Electronic Health Records
EHR-RAGp is a retrieval-augmented EHR foundation model that employs prototype-guided retrieval to dynamically integrate relevant historical patient context, outperforming prior models on clinical prediction tasks.
-
CAST: Collapse-Aware multi-Scale Topology Fusion for Multimodal Coreset Selection
CAST selects better multimodal coresets by fusing collapse-aware topologies across modalities and matching distributions at multiple scales in the diffusion wavelet domain.
-
BenchHAR: Benchmarking Self-Supervised Learning for Generalizable Sensor-based Activity Recognition
BenchHAR finds that hybrid reconstruction-plus-contrastive SSL with CNN encoders generalizes best for sensor HAR but overall performance on unseen distributions remains unsatisfactory.
-
Video Generation with Predictive Latents
PV-VAE improves video latent spaces for generation by unifying reconstruction with future-frame prediction, reporting 52% faster convergence and 34.42 FVD gain over Wan2.2 VAE on UCF101.
-
LLM4Log: A Systematic Review of Large Language Model-based Log Analysis
Systematic review of 145 papers on LLM-based log analysis, providing a unified taxonomy, common design patterns, evaluation practices, and challenges for deployment under drift and limited labels.
-
State Space Models for Bioacoustics: A Comparative Evaluation with Transformers
BioMamba matches Transformer performance on bioacoustics tasks while using significantly less VRAM.