Same Probe, Different Numbers: Are Activation Probes Robust to Inference-Time Numerical Non-Determinism?

arXiv:2609.31796 · cs.LG, cs.AI · Submitted 2026-09-25 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-09-25

Updated: 2026-09-25

Code: https://github.com/huggingface/transformers

Terminology

• Accuracy — 34× in this paper · explained in 1 episode
• Decode — 34× in this paper · explained in 2 episodes
• Transformer — 14× in this paper · explained in 2 episodes
• Alignment — 4× in this paper · explained in 1 episode
• Robustness — 4× in this paper · explained in 5 episodes
• AUROC — 3× in this paper · explained in 2 episodes
• Adam — 3× in this paper · explained in 2 episodes
• Cosine Similarity — 2× in this paper · explained in 3 episodes
• In Conference on Language Modeling — 2× in this paper
• Interpretability — 2× in this paper · explained in 15 episodes
• Quantization — 2× in this paper · explained in 12 episodes
• Benchmark — 1× in this paper · explained in 2 episodes
• Clustering — 1× in this paper · explained in 2 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• Hallucination — 1× in this paper · explained in 3 episodes
• International Conference on Machine Learning — 1× in this paper
• Intrinsic Dimension — 1× in this paper · explained in 2 episodes
• KV cache — 1× in this paper · explained in 3 episodes
• Linear probe — 1× in this paper · explained in 2 episodes
• Systematic Bias — 1× in this paper · explained in 3 episodes
• Trace — 1× in this paper · explained in 2 episodes
• Transfer Learning — 1× in this paper · explained in 10 episodes

Sources

• Understanding intermediate layers using linear classifier probes
• Non-Determinism of "Deterministic" LLM Settings
• Quantizable Transformers: Removing Outliers by Helping Attention Heads Do Nothing
• LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale
• Gemma 3 Technical Report
• Detecting Strategic Deception Using Linear Probes
• The Llama 3 Herd of Models
• Language Models Represent Space and Time
• Designing and Interpreting Probes with Control Tasks
• A Study of BFLOAT16 for Deep Learning Training
• Do All LLMs Know When They're Being Harmful? A Reproducibility Study of Latent-Space Safety Probes Across Model Families
• Similarity of Neural Network Representations Revisited
• BERT Busters: Outlier Dimensions that Disrupt Transformers
• Building Production-Ready Probes For Gemini
• The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets
• Detecting High-Stakes Interactions with Activation Probes
• Mixed Precision Training
• LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinations
• The Linear Representation Hypothesis and the Geometry of Large Language Models
• PyTorch: An Imperative Style, High-Performance Deep Learning Library

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks