Semantic Alignment of AI Models: Concept Collapse, Checkpoint Dynamics, and Cross-Lingual Transfer

arXiv:2608.01585 · cs.CL, cs.LG · Submitted 2026-08-03 · Read on arXiv

Tyler Ashoff, Jordan Rodu

cs.CL, cs.LG

Submitted: 2026-08-03

Comments: Code available at github.com/tylerashoff/persiscope (PyPI: persiscope)

Code: https://github.com/tylerashoff/persiscope

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Embeddings — 23× in this paper · explained in 2 episodes
• Benchmark — 17× in this paper · explained in 2 episodes
• Transformer — 14× in this paper · explained in 2 episodes
• Wasserstein Distance — 11× in this paper · explained in 6 episodes
• Papers — 5× in this paper · explained in 3 episodes
• AI Alignment — 3× in this paper · explained in 2 episodes
• Benchmarking — 3× in this paper · explained in 4 episodes
• Failure Modes — 3× in this paper · explained in 3 episodes
• Robustness — 3× in this paper · explained in 6 episodes
• Transformers — 3× in this paper · explained in 2 episodes
• Calibration — 2× in this paper · explained in 9 episodes
• Clustering — 2× in this paper · explained in 2 episodes
• Cross-lingual Transfer — 2× in this paper · explained in 2 episodes
• Decoder — 2× in this paper · explained in 2 episodes
• Equivariance — 2× in this paper · explained in 3 episodes
• Generalization — 2× in this paper · explained in 14 episodes
• Generative Model — 2× in this paper · explained in 2 episodes
• Generative Models — 2× in this paper · explained in 3 episodes
• Holistic Evaluation of Language Models — 2× in this paper
• Interpretability — 2× in this paper · explained in 15 episodes
• Knowledge Graph — 2× in this paper · explained in 8 episodes
• Knowledge graphs — 2× in this paper · explained in 2 episodes
• Latent Space — 2× in this paper · explained in 7 episodes
• Measuring Massive Multitask Language Understanding — 2× in this paper
• Mechanistic Interpretability — 2× in this paper · explained in 4 episodes
• Representation Engineering — 2× in this paper · explained in 2 episodes
• Representation Learning — 2× in this paper · explained in 3 episodes
• Semantic Alignment — 2× in this paper · explained in 6 episodes
• Agentic Reasoning — 1× in this paper · explained in 2 episodes
• Coherence — 1× in this paper · explained in 2 episodes
• Consistency — 1× in this paper · explained in 3 episodes
• Core — 1× in this paper · explained in 2 episodes
• Distillation — 1× in this paper · explained in 2 episodes
• Domain Shift — 1× in this paper · explained in 7 episodes
• Editor — 1× in this paper · explained in 2 episodes
• Evaluation Framework — 1× in this paper · explained in 3 episodes
• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 8 episodes
• Homogeneity — 1× in this paper · explained in 2 episodes
• Indirect Prompt Injection — 1× in this paper · explained in 2 episodes
• Knowledge Distillation — 1× in this paper · explained in 15 episodes
• Language Models (LLMs) — 1× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 9 episodes
• Latent Representation — 1× in this paper · explained in 2 episodes
• Latent Representations — 1× in this paper · explained in 3 episodes
• LoRA Adaptation — 1× in this paper · explained in 2 episodes
• Pairwise Comparisons — 1× in this paper · explained in 2 episodes
• Prompt Engineering — 1× in this paper · explained in 6 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Reward Hacking — 1× in this paper · explained in 11 episodes
• Semantic Understanding — 1× in this paper · explained in 3 episodes
• Sparse Autoencoders (SAEs) — 1× in this paper · explained in 4 episodes
• Tokenization — 1× in this paper · explained in 2 episodes
• Transformer Model — 1× in this paper · explained in 2 episodes
• Transformer Models — 1× in this paper · explained in 3 episodes

Sources

• Concrete Problems in AI Safety
• Constitutional AI: Harmlessness from AI Feedback
• Demonstrating specification gaming in reasoning models
• Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision
• ARC Prize 2025: Technical Report
• No Language Left Behind: Scaling Human-Centered Machine Translation
• Sparse Autoencoders Find Highly Interpretable Features in Language Models
• TranslateGemma Technical Report
• Alignment faking in large language models
• Language Models Represent Space and Time
• Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
• AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models
• The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets
• Efficient Estimation of Word Representations in Vector Space
• Emergent Linear Representations in World Models of Self-Supervised Sequence Models
• 2 OLMo 2 Furious
• The Geometry of Categorical and Hierarchical Concepts in Large Language Models
• Multilingual E5 Text Embeddings: A Technical Report
• Representation Engineering: A Top-Down Approach to AI Transparency
• Universal and Transferable Adversarial Attacks on Aligned Language Models

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering