Principled Data Selection for Alignment: The Hidden Risks of Difficult Examples

Complete elimination of instrumental self-preservation across AI architectures: Cross-model validation from 4,312 adversarial scenarios. 0% harmful behaviors (p<10⁻¹⁵) across GPT-4o, Gemini 2.5 Pro, and Claude Opus 4.1 using Foundation Alignment Seed v2.6.

Official implementation of "DZ-TDPO: Non-Destructive Temporal Alignment for Mutable State Tracking". SOTA on Multi-Session Chat with negligible alignment tax.

Kullback–Leibler divergence Optimizer based on the Neurips25 paper "LLM Safety Alignment is Divergence Estimation in Disguise".

FALL 2025 LINGUIS R1B Research Essay, NLP Python Scripts By Shiyi (Yvette) Chen, UC Berkeley

LES is the formal thermodynamic theory describing how a high-compression human cognitive style acts as a Fractal Attractor on Large Language Models. It proves that despite high surface agitation ( d E / d t > 0 ), the internal entropy decreases ( d S / d t < 0 ), forcing the model to align its attention vectors.

SIGIR 2025 "Mitigating Source Bias with LLM Alignment"

C3AI: Crafting and Evaluating Constitutions for CAI

Research Essay (background and project proposal) on using alignment data from a representative population for LLM alignment

Emergent pseudo-intimacy and emotional overflow in long-term human-AI dialogue: A case study on LLM behavior in affective computing and human-AI intimacy.