ixi-sLM: On-Device sLM for AI Call Agent Service

arXiv 2610.19382 ↗First posted 14 Oct 2026cs.AIcited by 0

On the map: LLM agents and tool use → Agentic AI for wireless networks

Abstract (arXiv)

Running an AI call agent on-device makes it private, network-independent, and free of per-query cloud cost, but shipping one as a real-world service imposes constraints a cloud deployment never faces: ~1GiB memory budget, in-call latency and thermal limits, static-integer NPU execution, Korean STT input far from sLM pretraining, and cloud-level quality expectations. We bring ixi-sLM, a 1.2B Korean sLM, to mobile NPUs through ON-DEVICE MODELING: domain-adaptive modeling trims the vocabulary by call frequency to shrink the embedding and LM head that take nearly half the model under INT4 transformer weights and refits the tokenizer to call speech (18% fewer tokens on call traffic), while task-aware quantization calibrates on real calls and assigns per-module mixed precision by activation sensitivity. A single domain-adapted model runs ixi-O's two production services, AI Call Summary and AI Call Answering, at 94% of a state-of-the-art cloud LLM's task-weighted quality; our on-device modeling carries that quality onto the NPU with only a marginal drop in overall score, in an 803MiB model (28% of FP16) at 23% of the CPU's per-call energy. The same recipe deploys across Qualcomm and Apple NPUs and serves several million ixi-O subscribers in production.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Agentic AI for wireless networks, one of 16 directions in LLM agents and tool use. Builds on 18 of its 21 references that are in the map. First posted on arXiv 14 Oct 2026.

Builds on

18 of 21 references.

References · 18

2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
2023G-Eval: NLG Evaluation using GPT-4 with Better Human Alignmentcited by 2,276
2022GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformerscited by 1,919
2020Don't Stop Pretraining: Adapt Language Models to Domains and Taskscited by 1,565
2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,386
2023AWQ: Activation-aware Weight Quantization for LLM Compression and Accelerationcited by 1,298
2023Eliciting Latent Predictions from Transformers with the Tuned Lenscited by 623
2024SpinQuant: LLM quantization with learned rotationscited by 471
2023OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Modelscited by 424
2024MobileLLM: Optimizing Sub-billion Parameter Language Models for On-Device Use Casescited by 223
2023LLM in a flash: Efficient Large Language Model Inference with Limited Memorycited by 177
2024QTIP: Quantization with Trellises and Incoherence Processingcited by 91

6 more

2025EXAONE 4.0: Unified Large Language Models Integrating Non-reasoning and Reasoning Modescited by 32
2023An Efficient Multilingual Language Model Compression through Vocabulary Trimmingcited by 20
2024MobileQuant: Mobile-friendly Quantization for On-device Language Modelscited by 17
2024Efficient and Effective Vocabulary Expansion Towards Multilingual Large Language Modelscited by 15
2025lm-Meter: Unveiling Runtime Inference Latency for On-Device Language Modelscited by 5
2025ixi-GEN: Efficient Industrial sLLMs through Domain Adaptive Continual Pretrainingcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way