DeEscalWild: A Real-World Benchmark for Automated De-Escalation Training with SLMs

• URL: http://arxiv.org/abs/2604.13075v1
• Date: Fri, 20 Mar 2026 23:32:28 GMT
• Title: DeEscalWild: A Real-World Benchmark for Automated De-Escalation Training with SLMs
• Abstract summary: Large Language Models (LLMs) enable dynamic, open-ended simulations, their substantial computational footprint renders them impractical for immersive field training.<n>Small Language Models (SLMs) offer a viable real-time alternative but suffer from a critical scarcity of high-quality, domain-specific training data.<n>We present DeEscalWild, a novel benchmark dataset curated from a pipeline of in-the-wild police-civilian interactions extracted from open-source video repositories.
• Score: 0.07249400282852116
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Effective de-escalation is critical for law enforcement safety and community trust, yet traditional training methods lack scalability and realism. While Large Language Models (LLMs) enable dynamic, open-ended simulations, their substantial computational footprint renders them impractical for deployment on the lightweight, portable hardware required for immersive field training. Small Language Models (SLMs) offer a viable real-time alternative but suffer from a critical scarcity of high-quality, domain-specific training data. To bridge this gap, we present DeEscalWild, a novel benchmark dataset curated from a multi-stage pipeline of in-the-wild police-civilian interactions extracted from open-source video repositories. Starting with 5,000 raw inputs, we employed a rigorous hybrid filtering process - combining human-in-the-loop verification with LLM-as-a-Judge evaluation - to distill 1,500 high-fidelity scenarios. The resulting corpus comprises 285,887 dialogue turns, totaling approximately 4.7 million tokens. Extensive experiments demonstrate that SLMs fine-tuned on this data significantly outperform their base counterparts across ROUGE-L, BLEU-4, METEOR, and BERTScore metrics. Notably, our fine-tuned Qwen 2.5 (3B-Instruct) surpasses the general-purpose Gemini 2.5 Flash model, demonstrating that domain-optimized SLMs can achieve superior performance with a fraction of the computational cost. This work establishes the foundational infrastructure for accessible, low-latency, and privacy-preserving officer training systems at the edge.

Related papers

• Proving the Utility of Large Language Models in Cybersecurity Simulations: A Comprehensive Examination [4.208276864195654]
  This paper explores how Large Language Models (LLMs) can bolster cybersecurity simulations.<n>We employ YAML as a structured representation format for simulating complex network configurations.<n> Comparative studies examine the advantages of LLM-based techniques over classical approaches.
  arXiv  Detail & Related papers  (2026-08-17T11:22:20Z)
• A Lightweight Defense Mechanism against Next Generation of Phishing Emails using Distilled Attention-Augmented BiLSTM [34.0814379994364]
  The MobileBERT teacher receives fine-tuning before its transformation into a BiLSTM model with multi-head attention.<n>The system demonstrates excellent performance in terms of accuracy and latency while maintaining a compact size.<n>The paper examines system performance under high traffic conditions and security measures for privacy protection and implementation methods for operational deployment.
  arXiv  Detail & Related papers  (2026-02-24T20:06:45Z)
• Language-based Trial and Error Falls Behind in the Era of Experience [50.503828360874536]
  Large Language Models (LLMs) excel in language-based agentic tasks, but their applicability to unseen, nonlinguistic environments remains limited.<n>In this work, we demonstrate the primary bottleneck is the prohibitive cost of exploration.<n>We propose SCOUT, a novel framework that decouples exploration from semantic exploitation.
  arXiv  Detail & Related papers  (2026-01-29T14:08:41Z)
• Towards High Data Efficiency in Reinforcement Learning with Verifiable Reward [54.708851958671794]
  We propose a Data-Efficient Policy Optimization pipeline that combines optimized strategies for both offline and online data selection.<n>In offline phase, we curate a high-quality subset of training samples based on diversity, influence, and appropriate difficulty.<n>During online RLVR training, we introduce a sample-level explorability metric to dynamically filter samples with low exploration potential.
  arXiv  Detail & Related papers  (2025-09-01T10:04:20Z)
• Pentest-R1: Towards Autonomous Penetration Testing Reasoning Optimized via Two-Stage Reinforcement Learning [12.155067662797281]
  Pentest-R1 is a framework designed to optimize reasoning capabilities for penetration testing tasks.<n>It learns directly from environmental feedback to develop robust error self-correction and adaptive strategies.<n>On AutoPenBench, Pentest-R1 achieves a 24.2% success rate, surpassing most state-of-the-art models.
  arXiv  Detail & Related papers  (2025-08-10T15:14:05Z)
• Filter-And-Refine: A MLLM Based Cascade System for Industrial-Scale Video Content Moderation [21.18948097612397]
  This paper introduces an efficient method to transform a generative MLLM into a multimodal classifier using minimal discriminative training data.<n>We then propose a router-ranking cascade system that integrates MLLMs with a lightweight router model.<n>Online evaluations show that our system increases automatic content moderation volume by 41%, while the cascading deployment reduces computational cost to only 1.5% of direct full-scale deployment.
  arXiv  Detail & Related papers  (2025-07-23T04:52:58Z)
• SHA256 at SemEval-2025 Task 4: Selective Amnesia -- Constrained Unlearning for Large Language Models via Knowledge Isolation [12.838593066237452]
  Large language models (LLMs) memorize frequently sensitive information during training, posing risks when deploying publicly accessible models.<n>This paper presents our solution to SemEval-2025 Task 4 on targeted unlearning, which combines causal mediation analysis with layer-specific optimization.
  arXiv  Detail & Related papers  (2025-04-17T15:05:40Z)
• Synthesizing Post-Training Data for LLMs through Multi-Agent Simulation [51.20656279478878]
  MATRIX is a multi-agent simulator that automatically generates diverse text-based scenarios.<n>We introduce MATRIX-Gen for controllable and highly realistic data synthesis.<n>On AlpacaEval 2 and Arena-Hard benchmarks, Llama-3-8B-Base, post-trained on datasets synthesized by MATRIX-Gen with just 20K instruction-response pairs, outperforms Meta's Llama-3-8B-Instruct model.
  arXiv  Detail & Related papers  (2024-10-18T08:01:39Z)
• InternLM2 Technical Report [159.70692271378581]
  This paper introduces InternLM2, an open-source Large Language Models (LLMs) that outperforms its predecessors in comprehensive evaluations across 6 dimensions and 30 benchmarks. The pre-training process of InternLM2 is meticulously detailed, highlighting the preparation of diverse data types. InternLM2 efficiently captures long-term dependencies, initially trained on 4k tokens before advancing to 32k tokens in pre-training and fine-tuning stages.
  arXiv  Detail & Related papers  (2024-03-26T00:53:24Z)
• MobiLlama: Towards Accurate and Lightweight Fully Transparent GPT [87.4910758026772]
  "Bigger the better" has been the predominant trend in recent Large Language Models (LLMs) development. This paper explores the "less is more" paradigm by addressing the challenge of designing accurate yet efficient Small Language Models (SLMs) for resource constrained devices.
  arXiv  Detail & Related papers  (2024-02-26T18:59:03Z)
• FedDBL: Communication and Data Efficient Federated Deep-Broad Learning for Histopathological Tissue Classification [65.7405397206767]
  We propose Federated Deep-Broad Learning (FedDBL) to achieve superior classification performance with limited training samples and only one-round communication. FedDBL greatly outperforms the competitors with only one-round communication and limited training samples, while it even achieves comparable performance with the ones under multiple-round communications. Since no data or deep model sharing across different clients, the privacy issue is well-solved and the model security is guaranteed with no model inversion attack risk.
  arXiv  Detail & Related papers  (2023-02-24T14:27:41Z)

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.