mlsanityData to 7 Oct 2026

NILE: Internal Consistency Alignment in Large Language Models

arXiv 2412.16686 ↗First posted 21 Dec 2024cs.CLcited by 2

Not on the map. Why?

Abstract (arXiv)

As a crucial step to enhance LLMs alignment with human intentions, Instruction Fine-Tuning (IFT) has a high demand on dataset quality. However, existing IFT datasets often contain knowledge that is inconsistent with LLMs' internal knowledge learned from the pre-training phase, which can greatly affect the efficacy of IFT. To address this issue, we introduce NILE (iNternal consIstency aLignmEnt) framework, aimed at optimizing IFT datasets to unlock LLMs' capability further. NILE operates by eliciting target pre-trained LLM's internal knowledge corresponding to instruction data. The internal knowledge is leveraged to revise the answer in IFT datasets. Additionally, we propose a novel Internal Consistency Filtering (ICF) method to filter training samples, ensuring its high consistency with LLM's internal knowledge. Our experiments demonstrate that NILE-aligned IFT datasets sharply boost LLM performance across multiple LLM ability evaluation datasets, achieving up to 66.6% gain on Arena-Hard and 68.5% on Alpaca-Eval V2. Further analysis confirms that each component of the NILE}framework contributes to these substantial performance improvements, and provides compelling evidence that dataset consistency with pre-trained internal knowledge is pivotal for maximizing LLM potential.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 32 of its 49 references that are in the map. First posted on arXiv 21 Dec 2024.

Builds on

32 of 49 references.

References · 32

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  4. 2022Self-Instruct: Aligning Language Models with Self-Generated Instructionscited by 2,674
  5. 2022Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language modelscited by 1,970
  6. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,767
  7. 2018Know What You Don't Know: Unanswerable Questions for SQuADcited by 1,538
  8. 2024Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preferencecited by 1,151
  9. 2023LIMA: Less Is More for Alignmentcited by 1,080
  10. 2024Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluatorscited by 865
  11. 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 858
  12. 2023Enhancing Chat Language Models by Scaling High-quality Instructional Conversationscited by 825
20 more
  1. 2023The Flan Collection: Designing Data and Methods for Effective Instruction Tuningcited by 734
  2. 2024Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Modelscited by 552
  3. 2024From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipelinecited by 484
  4. 2022Unnatural Instructions: Tuning Language Models with (Almost) No Human Laborcited by 395
  5. 2024Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothingcited by 341
  6. 2023Principle-Driven Self-Alignment of Language Models from Scratch with Minimal Human Supervisioncited by 318
  7. 2023From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuningcited by 301
  8. 2023Self-Alignment with Instruction Backtranslationcited by 149
  9. 2024Unfamiliar Finetuning Examples Control How Language Models Hallucinatecited by 85
  10. 2023Cue-CoT: Chain-of-thought Prompting for Responding to In-depth Dialogue Questions with LLMscited by 62
  11. 2024Key-Point-Driven Data Synthesis with its Enhancement on Mathematical Reasoningcited by 58
  12. 2023Instruction Mining: Instruction Data Selection for Tuning Large Language Modelscited by 57
  13. 2024Selective Reflection-Tuning: Student-Selected Data Recycling for LLM Instruction-Tuningcited by 57
  14. 2024Learning or Self-aligning? Rethinking Instruction Fine-tuningcited by 43
  15. 2024Image Textualization: An Automatic Framework for Creating Accurate and Detailed Image Descriptionscited by 43
  16. 2023Unsupervised Dense Retrieval with Relevance-Aware Contrastive Pre-Trainingcited by 35
  17. 2023Enhancing Large Language Models Against Inductive Instructions with Dual-critique Promptingcited by 21
  18. 2023CoachLM: Automatic Instruction Revisions Improve the Data Quality in LLM Instruction Tuningcited by 19
  19. 2024Self-Tuning: Instructing LLMs to Effectively Acquire New Knowledge through Self-Teachingcited by 17
  20. 2024Case2Code: Scalable Synthetic Data for Code Generationcited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025Self-Reasoning Language Models: Unfold Hidden Reasoning Chains with Few Reasoning Catalystcited by 6
  2. 2026Synthesizing Instruction-Tuning Datasets with Contrastive Decodingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Enhancing and Assessing Instruction-Following with Fine-Grained Instruction Variants
  2. 2025Towards Alignment-Centric Paradigm: A Survey of Instruction Tuning in Large Language Models
  3. 2024Long Is More for Alignment: A Simple but Tough-to-Beat Baseline for Instruction Fine-Tuning
  4. 2024Improving the Robustness of Large Language Models via Consistency Alignment
  5. 2025A Comprehensive Evaluation framework of Alignment Techniques for LLMs
  6. 2023The Unlocking Spell on Base LLMs: Rethinking Alignment via In-Context Learning
  7. 2025Massive Supervised Fine-tuning Experiments Reveal How Data, Layer, and Training Factors Shape LLM Alignment Quality
  8. 2025Condor: Enhance LLM Alignment with Knowledge-Driven Data Synthesis and Refinement