ECLIPTICA -- A Framework for Switchable LLM Alignment via CITA - Contrastive Instruction-Tuned Alignment

arXiv:2601.06157 · cs.LG, cs.AI · Submitted 2026-01-06 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-01-06

Updated: 2026-10-11

Code: https://github.com/acl-org/aclpubc

Terminology

• Alignment — 9× in this paper · explained in 1 episode
• Contrastive Instruction-Tuned Alignment — 2× in this paper
• GRPO — 2× in this paper · explained in 6 episodes
• Benchmark — 1× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes

Sources

• A General Theoretical Paradigm to Understand Learning from Human Preferences
• Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
• Constitutional AI: Harmlessness from AI Feedback
• Are aligned neural networks adversarially aligned?
• Scaling Instruction-Finetuned Language Models
• The Llama 3 Herd of Models
• KTO: Model Alignment as Prospect Theoretic Optimization
• Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned
• Co-Pilot for Health: Personalized Algorithmic AI Nudging to Improve Health Outcomes
• ORPO: Monolithic Preference Optimization without Reference Model
• Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation
• OPT-IML: Scaling Language Model Instruction Meta Learning through the Lens of Generalization
• PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference
• Mistral 7B
• Holistic Evaluation of Language Models
• Diffusion Language Models Are Versatile Protein Learners
• Jailbreaking ChatGPT via Prompt Engineering: An Empirical Study
• SimPO: Simple Preference Optimization with a Reference-Free Reward
• Orca: Progressive Learning from Complex Explanation Traces of GPT-4
• GPT-4 Technical Report

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: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks