REVIEW 20 cited by
Small Language Models: Survey, Measurements, and Insights
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
Small Language Models: Survey, Measurements, and Insights
read the original abstract
Small language models (SLMs), despite their widespread adoption in modern smart devices, have received significantly less academic attention compared to their large language model (LLM) counterparts, which are predominantly deployed in data centers and cloud environments. While researchers continue to improve the capabilities of LLMs in the pursuit of artificial general intelligence, SLM research aims to make machine intelligence more accessible, affordable, and efficient for everyday tasks. Focusing on transformer-based, decoder-only language models with 100M-5B parameters, we survey 70 state-of-the-art open-source SLMs, analyzing their technical innovations across three axes: architectures, training datasets, and training algorithms. In addition, we evaluate their capabilities in various domains, including commonsense reasoning, mathematics, in-context learning, and long context. To gain further insight into their on-device runtime costs, we benchmark their inference latency and memory footprints. Through in-depth analysis of our benchmarking data, we offer valuable insights to advance research in this field.
Forward citations
Cited by 20 Pith papers
-
Does Mixture-of-Experts Actually Help Inference on Consumer and Edge Hardware? An Empirical Study
On bandwidth-bound edge hardware, MoE inference cost tracks total parameters rather than active ones, so sparse activation fails to deliver the expected throughput or energy gains.
-
Offline Vision-Language Navigation with Geometric Goal Localization for Outdoor Environments
An offline vision-language navigation system matches cloud-based performance using a small onboard language model and LiDAR-grounded goal localization, succeeding in 31/32 outdoor trials.
-
Small, Free, and Effective: Orchestrating Open-Weight Small Language Models to Outperform Single LLM for Malware Analysis
An evidence-grounded hybrid of 4B and 8B open-weight models reaches 35.30% on malware-report QA, slightly beating the best ungrounded frontier model (34.77%) while trailing a frontier model given the same evidence (38.22%).
-
In-Context Learning for Wound Classification with Small Multimodal Language Models
Retrieval-based in-context learning, not zero-shot prompting, drives wound-classification gains in small multimodal models, with Qwen 3.5 27B reaching 0.872 accuracy on Kaggle and 0.678 on Medetec.
-
Punching Above Their Weight: Classification-Head Fine-Tuning of Tiny Language Models (TLMs) for Verifiable Multiple-Choice Tasks
Classification-head LoRA fine-tuning of sub-3B Qwen3 models outperforms label-generation SFT by 2–3% on HellaSwag, WinoGrande and PIQA and yields SOTA numbers competitive with GPT-3/PaLM/GPT-4.
-
To Isolate or to Score? Model-Adaptive Assessment for Cost-Efficient Multi-Agent RAG
Empirical study finds isolation drives gains for weak models in multi-agent RAG while scoring matters for strong ones, enabling MADARA for cost-efficient adaptive assessment.
-
AgentFloor: How Far Up the tool use Ladder Can Small Open-Weight Models Go?
Small open-weight models match GPT-5 on routine agent tool-use tasks but lag on long-horizon planning, supporting tiered routing to reduce costs in agentic systems.
-
Local-Splitter: A Measurement Study of Seven Tactics for Reducing Cloud LLM Token Usage on Coding-Agent Workloads
Combining local routing with prompt compression saves 45-79% cloud tokens on edit and explanation workloads, while a fuller set including draft-review saves 51% on RAG-heavy tasks.
-
EdgeCIM: A Hardware-Software Co-Design for CIM-Based Acceleration of Small Language Models
A CIM-based hardware-software co-design in 65nm achieves up to 7.3x higher throughput and 49.59x better energy efficiency than NVIDIA Orin Nano for LLaMA3.2-1B, averaging 336 tokens/s and 173 tokens/J under INT4 acros...
-
Taming Asynchronous CPU-GPU Coupling for Frequency-aware Latency Estimation on Mobile Edge
FLAME models layer-wise overlapping parallelism and asynchronous CPU-GPU pipeline bubbles to estimate inference latency across frequencies with sparse profiling and low error for DNNs and SLMs.
-
MIITA: Memory-Induced Inference-Time Adaptation for Continual Learning with Small Language Models
MIITA stores past supervised examples as compact hidden-space correction directions and applies retrieved directions through gated temporary hidden-state updates, improving continual learning in small language models ...
-
Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs
EngGPT2MoE-16B-A3B matches or exceeds other Italian open-source LLMs on most international benchmarks while remaining competitive on ITALIC, though it trails some top international models.
-
Domain-Adapted Small Language Models for Reliable Clinical Triage
Domain-adapted Qwen2.5-7B small language models reduce discordance and clinically significant errors in ESI triage compared to baselines and larger LLMs.
-
Small Language Models are the Future of Agentic AI
Small language models are sufficiently capable, more suitable, and far more economical than large models for the repetitive tasks that dominate agentic AI systems.
-
Enhancing Small Language Models Reasoning through Knowledge Graph Grounding
On CLUTRR, adding RGCN-generated kinship hints improves small models by 1.1-1.7x over story-only baselines, but noisy self-extracted facts cap gains and sometimes hurt.
-
Does Mixture-of-Experts Actually Help Inference on Consumer and Edge Hardware? An Empirical Study
Empirical benchmarks show MoE inference cost on edge hardware tracks total parameters rather than active parameters, with OLMoE-1B-7B behind dense baselines especially on the Jetson device.
-
Agents Should Replace Narrow Predictive AI as the Orchestrator in 6G AI-RAN
Position paper proposes replacing fragmented narrow AI models with LLMs as the cognitive orchestrator in the RAN Intelligent Controller for Level 5 autonomous 6G networks.
-
Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs
EngGPT2MoE-16B-A3B matches or beats other Italian models on most international benchmarks but trails top international models such as GPT-5 nano and Qwen3-8B.
-
Sustainable Code Generation Using Large Language Models: A Systematic Literature Review
A systematic review finds research on the sustainability of LLM-generated code to be limited, fragmented, and without accepted frameworks for measurement or benchmarking.
-
Will LLMs Scaling Hit the Wall? Breaking Barriers via Distributed Resources on Massive Edge Devices
Position paper claiming that distributed training across massive edge devices can overcome data depletion and centralized compute monopolies in LLM scaling.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.