🔬 Mech Interp - taylor · Scour

A Unifying Framework for Concept-Based Representational Similarity

🔍Interpretability Academic

Inside the LLM Word Factory

🔧Compilers Academic

Priors Persist Through Suppression: A Stroop Paradigm for Lexical Override

🔣Category Theory Academic

What's in a Name? Morphological Shortcuts by LLMs in Pharmacology

🔧Compilers Academic

DiffoR: A Unified Continuous Generative Framework for Universal Ordinal Regression

🛡️AI Safety Academic

The Amplifying Mirror: Locating and Steering the Partisan Direction inside a Large Language Model

🔍Interpretability Academic

When Built-in Thinking Helps and Hurts: Constraint-Level Error Shifts in Instruction Following

🛡️AI Safety Academic

Beyond Text Following: Repairable Arbitration Reversals in Audio-Language Models

🌐Distributed Systems Academic

Mechanistic Insights into Functional Sparsity in Multimodal LLMs via CoRe Heads

🔍Interpretability Academic

The Neutral Mask: How RLHF Provides Shallow Alignment while Leaving Partisan Structure Intact in a Large Language Model

🛡️AI Safety Academic

Wearable Single-Lead ECG Detects Fine-Grained Structural Heart Disease Through Echo-Report Supervision

🔍Interpretability Academic

Arithmetic Pedagogy for Language Models

🔍Interpretability Academic

arxiv.org··Hacker News

Unsupervised Pattern Analysis in Japanese Veterinary Toxicology: A Regulatory-Compliant Framework for Cross-Species Risk Assessment

🛡️AI Safety Academic

Log in to enable infinite scrolling