Pith. sign in

REVIEW 7 cited by

Beyond IID: Optimizing Instruction Learning from the Perspective of Instruction Interaction and Dependency

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2409.07045 v1 pith:C5KH33NV submitted 2024-09-11 cs.CL cs.AI

classification cs.CLcs.AI
keywords instructioninstructionsinteractiondependencydifferentlearningoptimizepatterns
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

With the availability of various instruction datasets, a pivotal challenge is how to effectively select and integrate these instructions to fine-tune large language models (LLMs). Previous research mainly focuses on selecting individual high-quality instructions. However, these works overlooked the joint interactions and dependencies between different categories of instructions, leading to suboptimal selection strategies. Moreover, the nature of these interaction patterns remains largely unexplored, let alone optimize the instruction set with regard to them. To fill these gaps, in this paper, we: (1) systemically investigate interaction and dependency patterns between different categories of instructions, (2) manage to optimize the instruction set concerning the interaction patterns using a linear programming-based method, and optimize the learning schema of SFT using an instruction dependency taxonomy guided curriculum learning. Experimental results across different LLMs demonstrate improved performance over strong baselines on widely adopted benchmarks.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. SPaR: Self-Play with Tree-Search Refinement to Improve Instruction-Following in Large Language Models

    cs.CL 2024-12 conditional novelty 7.0 of 10

    SPaR trains LLMs by having them refine their own failed responses through tree search, producing preference pairs with minimal irrelevant variation, and lifts LLaMA3-8B above GPT-4-Turbo on IFEval.

  2. Structural Entropy Guided Agent for Detecting and Repairing Knowledge Deficiencies in LLMs

    cs.CL 2025-05 conditional novelty 6.0 of 10

    SENATOR guides a language model through a knowledge graph, measures its uncertainty with structural entropy, and fine-tunes it on synthetic data chosen to fix its weak spots, gaining up to 12 percent average relative ...

  3. PersonaFuse: A Personality Activation-Driven Framework for Enhancing Human-LLM Interactions

    cs.CL 2025-09 reject novelty 5.0 of 10

    A post-training framework with persona-specific LoRA experts and a situation-aware router improves LLM emotional responses, but the evidence on preserving general ability is undercut by missing base-model comparisons.

  4. Data Mixing Optimization for Supervised Fine-Tuning of Large Language Models

    cs.AI 2025-08 unverdicted novelty 5.0 of 10

    The authors introduce and test an optimization method for fine-tuning data mixtures that matches costly grid-search results within 0.66% validation loss.

  5. Infinity Instruct: Scaling Instruction Selection and Synthesis to Enhance Language Models

    cs.CL 2025-06 conditional novelty 5.0 of 10

    A two-phase pipeline selects and synthesizes 8.9M instruction examples that, after fine-tuning, push open-source LLMs ahead of their official chat-tuned versions on both foundational and conversational benchmarks.

  6. DeepDistill: Enhancing LLM Reasoning Capabilities via Large-Scale Difficulty-Graded Data Training

    cs.CL 2025-04 conditional novelty 5.0 of 10

    A two-stage SFT recipe using pass-rate and coefficient-of-variation based data selection from a 40M-response distilled dataset lifts Qwen2.5-72B to 79.2% on AIME2024, nearly matching RL-trained reasoning models.

  7. Exploring the Potential of Offline RL for Reasoning in LLMs: A Preliminary Study

    cs.CL 2025-05 conditional novelty 3.0 of 10

    Applying LD-DPO to the DeepDistill-32B model improves average benchmark scores by 3.3 points, but the gains are uneven, the tables are inconsistent, and the evidence is preliminary.

Pith tools