LLM Ghostbusters: Surgical Hallucination Suppression via Adaptive Unlearning

• URL: http://arxiv.org/abs/2605.01047v1
• Date: Fri, 01 May 2026 19:20:31 GMT
• Title: LLM Ghostbusters: Surgical Hallucination Suppression via Adaptive Unlearning
• Abstract summary: We present Adaptive Unlearning, a post-deployment framework that surgically suppresses hallucinations while preserving general model utility.<n>We demonstrate that AU reduces package rates by 81%, corresponding to a substantial reduction in slopsquatting attack surface.<n>Our analysis shows that distributional changes are concentrated on package-related generations, leaving general coding behavior largely unaffected.
• Score: 12.855537727854975
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Hallucinations, outputs that sound plausible but are factually incorrect, remain an open challenge for deployed LLMs. In code generation, models frequently hallucinate non-existent software packages, recommending imports and installation commands for fictional libraries. This creates a critical supply-chain vulnerability: an attacker can proactively register such packages on public registries with malicious payloads that are subsequently installed and executed by developers or autonomous agents, a class of package confusion attack known as slopsquatting. Once a model is deployed, mitigating this failure mode is difficult: full retraining is costly, and existing approaches either cause severe degradation of model utility or rely on a pre-specified forget-set, an assumption that does not apply to the unbounded space of hallucinations. To address this problem, we present Adaptive Unlearning (AU), a post-deployment framework that surgically suppresses hallucinations while preserving general model utility. AU introduces a hybrid token-level objective that simultaneously reinforces valid outputs and suppresses hallucinated ones. Combined with an adaptive discovery loop that continuously surfaces new hallucination-inducing contexts without human supervision, AU enables generalization to unseen prompts and hallucinations. We demonstrate that AU reduces package hallucination rates by 81%, corresponding to a substantial reduction in slopsquatting attack surface, while maintaining performance on standard coding benchmarks. Our analysis shows that distributional changes are concentrated on package-related generations, leaving general coding behavior largely unaffected and confirming that AU's effect is isolated to the targeted distribution. AU operates entirely on model-generated data, requires no human annotation, and generalizes across domains.

Related papers

• Harmless Yet Harmful: Neutral Prompting Attacks for Stealthy Hallucination Steering in Agent Skills [15.148439025792017]
  This paper introduces emphNeutral Prompting Attack (NPA)<n>NPA is a stealthy attack paradigm in which semantically benign instructions, such as encouraging imagination, increase package hallucination propensity without containing explicit malicious intent.<n>Our results show that NPA increases both emphHallucination ASR and emphPip Install ASR, changes the distribution of hallucinated package names, and evades existing static-analysis, LLM-based, and agent-based Skill defenses.
  arXiv  Detail & Related papers  (2026-05-28T04:48:45Z)
• VCE: A zero-cost hallucination mitigation method of LVLMs via visual contrastive editing [70.82867621856968]
  Large vision-language models (LVLMs) frequently suffer from Object Hallucination (OH)<n>Recent studies suggest that the hallucination problem may stem from language priors.<n>We propose Visual Contrastive Editing (VCE), a novel post-hoc method that identifies and suppresses hallucinatory tendencies.
  arXiv  Detail & Related papers  (2026-04-21T12:40:07Z)
• Look Carefully: Adaptive Visual Reinforcements in Multimodal Large Language Models for Hallucination Mitigation [51.743225614196774]
  Multimodal large language models (MLLMs) have achieved remarkable progress in vision-language reasoning.<n>They remain vulnerable to hallucination, where generated content deviates from visual evidence.<n>Recent vision enhancement methods attempt to address this issue by reinforcing visual tokens during decoding.<n>We propose Adaptive Visual Reinforcement (AIR), a training-free framework for MLLMs.
  arXiv  Detail & Related papers  (2026-02-27T14:18:51Z)
• PruneHal: Reducing Hallucinations in Multi-modal Large Language Models through Adaptive KV Cache Pruning [87.35309934860938]
  hallucinations in large language models (MLLMs) are strongly associated with insufficient attention allocated to visual tokens.<n>We propose textbfPruneHal, a training-free, simple yet effective method that leverages adaptive KV cache pruning to enhance the model's focus on critical visual information.
  arXiv  Detail & Related papers  (2025-10-22T02:41:07Z)
• Mitigating Hallucinations in Large Vision-Language Models by Self-Injecting Hallucinations [73.37711261605271]
  hallucination mitigation methods are mainly based on preference alignment and require external human annotations or auxiliary models for preference data collection.<n>We propose Autonomous Preference Alignment via Self-Injection (APASI), a novel and generalizable method that mitigates hallucinations without external dependencies.<n>APASI leverages the target LVLM to self-inject hallucinations into a generated response, creating a pair of responses with varying preference levels.
  arXiv  Detail & Related papers  (2025-09-14T14:26:53Z)
• Osiris: A Lightweight Open-Source Hallucination Detection System [30.63248848082757]
  hallucinations prevent RAG systems from being deployed in production environments.<n>We introduce a multi-hop QA dataset with induced hallucinations.<n>We achieve better recall with a 7B model than GPT-4o on the RAGTruth hallucination detection benchmark.
  arXiv  Detail & Related papers  (2025-05-07T22:45:59Z)
• HalluLens: LLM Hallucination Benchmark [49.170128733508335]
  Large language models (LLMs) often generate responses that deviate from user input or training data, a phenomenon known as "hallucination"<n>This paper introduces a comprehensive hallucination benchmark, incorporating both new extrinsic and existing intrinsic evaluation tasks.
  arXiv  Detail & Related papers  (2025-04-24T13:40:27Z)
• Generate, but Verify: Reducing Hallucination in Vision-Language Models with Retrospective Resampling [78.78822033285938]
  Vision-Language Models (VLMs) excel at visual understanding but often suffer from visual hallucinations.<n>In this work, we introduce REVERSE, a unified framework that integrates hallucination-aware training with on-the-fly self-verification.
  arXiv  Detail & Related papers  (2025-04-17T17:59:22Z)
• Importing Phantoms: Measuring LLM Package Hallucination Vulnerabilities [11.868859925111561]
  Large Language Models (LLMs) have become an essential tool in the programmer's toolkit.<n>Their tendency to hallucinate code can be used by malicious actors to introduce vulnerabilities to broad swathes of the software supply chain.
  arXiv  Detail & Related papers  (2025-01-31T10:26:18Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.