Model Understanding
DV-Lens: Revealing the Functional Organization of Language Model Parameters
A vocabulary-space view of native parameters, tested through local output changes and targeted interventions.
Reliability & Safety
Do correct answers have stable latent knowledge support?
A correct answer is an observable outcome. This study asks whether its internal knowledge support stays stable when the elicitation condition changes.
Capability Engineering
SuperAI: Merging Capabilities Across Heterogeneous Foundation Models
Bring complementary capabilities into one host model by aligning functional subspaces across heterogeneous foundation models.
Applied Research
Forensic Reserve: Eliciting Latent Knowledge for Image Forgery Detection
Find a pretrained model’s latent forensic reserve, then constrain lightweight adaptation to those existing directions.
Reliability & Safety
EmoRSS: Mitigating Emotion-Induced Over-Refusal in Large Language Models
Emotional wording can make benign requests look unsafe. EmoRSS steers refusal-related activations to reduce that distortion.
Applied Research
ContractLens: Latent Security Knowledge for Malicious Smart Contract Detection
Localize latent security signals and use a compact set of internal units to detect malicious smart contracts.
Capability Engineering
BARQ: Balanced Codebook Refinement for Low-Bit LLM Quantization
Improve low-bit quantization by giving codewords balanced opportunities to learn during fitting.
Reliability & Safety
ACTR: Aligning Thoughts and Responses for Multilingual Safety in Reasoning LLMs
A reasoning model may recognize a safety risk in its thoughts yet still produce an unsafe answer in another language.
Capability Engineering
Hub-Spectral Activation of Latent Multimodal Knowledge
A shared hub can hide useful cross-modal relations that native similarity scores do not reliably expose.
Applied Research
Latent Anomaly Knowledge Excavation: Unveiling Sparse Sensitive Neurons in Vision-Language Models
Excavate anomaly-sensitive neurons already present in a vision-language model using a small set of normal samples.
Reliability & Safety
OrthoEraser: Coupled-Neuron Orthogonal Projection for Concept Erasure
Erase a sensitive concept while explicitly protecting the benign features coupled to the intervention.
Capability Engineering
Transport and Merge: Cross-Architecture Merging for Large Language Models
Use activation correspondences to transfer knowledge between language models whose architectures do not match.
Reliability & Safety
Who Transfers Safety? Identifying and Targeting Cross-Lingual Shared Safety Neurons
Identify shared safety neurons that connect high-resource and non-high-resource language behaviour.
Applied Research
DNA: Uncovering Universal Latent Forgery Knowledge
Look for forgery-sensitive anchors already encoded in pretrained models instead of relying only on end-to-end adaptation.
Reliability & Safety
TraceRouter: Robust Safety for Large Foundation Models via Path-Level Intervention
Follow the cross-layer propagation of harmful semantics and intervene at the pathway level.
Model Understanding
Where Culture Fades: Revealing the Cultural Gap in Text-to-Image Generation
A model may contain cultural knowledge yet fail to activate it when generating images from multilingual prompts.
