Pith. sign in

REVIEW 6 cited by

Fundamental Safety-Capability Trade-offs in Fine-tuning Large Language Models

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 2503.20807 v1 pith:GGIKJM6L submitted 2025-03-24 stat.ML cs.AIcs.CLcs.LG

classification stat.MLcs.AIcs.CLcs.LG
keywords fine-tuningsafety-capabilitybeencapabilityfundamentallanguagelargellms
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Fine-tuning Large Language Models (LLMs) on some task-specific datasets has been a primary use of LLMs. However, it has been empirically observed that this approach to enhancing capability inevitably compromises safety, a phenomenon also known as the safety-capability trade-off in LLM fine-tuning. This paper presents a theoretical framework for understanding the interplay between safety and capability in two primary safety-aware LLM fine-tuning strategies, providing new insights into the effects of data similarity, context overlap, and alignment loss landscape. Our theoretical results characterize the fundamental limits of the safety-capability trade-off in LLM fine-tuning, which are also validated by numerical experiments.

Discussion (0). Sign in to comment.

Forward citations

Cited by 6 Pith papers

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

  1. Helpfulness Hurts: Domain-Dependent Degradation of Mid-Trained Compassion Values Under Post-Training

    cs.CL 2026-04 unverdicted novelty 6.0 of 10

    Helpfulness SFT/GRPO degrades mid-trained animal compassion far more than coding SFT/GRPO on ANIMA, with the compassion effect transferring cross-lingually while general moral-reasoning gains do not.

  2. The Future is Agentic: Definitions, Perspectives, and Open Challenges of Multi-Agent Recommender Systems

    cs.IR 2025-07 conditional novelty 6.0 of 10

    A framework for agentic recommender systems plus a pilot study showing multi-agent pipelines beat a single-shot LLM only on high-diversity user histories.

  3. MGC: A Compiler Framework Exploiting Compositional Blindness in Aligned LLMs for Malware Generation

    cs.CR 2025-07 conditional novelty 6.0 of 10

    MGC, a two-stage compiler framework, generates functional malware by decomposing malicious intents into benign-appearing MDIR components that strong aligned LLMs will implement, bypassing safety alignment.

  4. Discovering Hierarchical Latent Capabilities of Language Models via Causal Representation Learning

    cs.LG 2025-06 reject novelty 6.0 of 10

    From Open LLM Leaderboard data grouped by base model, the authors recover a three-factor ordering of LLM capabilities and claim instruction-following causally supports math reasoning.

  5. SATORI: Static Test Oracle Generation for REST APIs

    cs.SE 2025-08 unverdicted novelty 5.0 of 10

    SATORI statically infers REST API test oracles from OpenAPI specs via LLMs, reporting F1 74.3%, above AGORA+'s 69.3%, with 18 confirmed bugs; the supplied full text, however, is a different paper.

  6. SEALGuard: Safeguarding the Multilingual Conversations in Southeast Asian Languages for LLM Software Systems

    cs.CL 2025-07 conditional novelty 5.0 of 10

    A LoRA-adapted SeaLLM model detects unsafe and jailbreak prompts in nine Southeast Asian languages with 97% recall and 98% F1 on the authors' new SEALSBench benchmark, far above zero-shot LlamaGuard.

Pith tools