Architectural Trade-offs in Small Language Models Under Compute Constraints

• URL: http://arxiv.org/abs/2512.20877v1
• Date: Wed, 24 Dec 2025 01:36:50 GMT
• Title: Architectural Trade-offs in Small Language Models Under Compute Constraints
• Authors: Shivraj Singh Bhatti,
• Abstract summary: We present a systematic study of small language models under strict compute constraints.<n>We evaluate each on character-level modeling of Tiny Shakespeare and word-level modeling of Penn Treebank (PTB) and WikiText-2.<n>Our results show that attention-based models dominate per-FLOP efficiency even at small scale, while increasing depth or context can degrade performance.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present a systematic empirical study of small language models under strict compute constraints, analyzing how architectural choices and training budget interact to determine performance. Starting from a linear next-token predictor, we progressively introduce nonlinearities, self-attention, and multi-layer transformer architectures, evaluating each on character-level modeling of Tiny Shakespeare and word-level modeling of Penn Treebank (PTB) and WikiText-2. We compare models using test negative log-likelihood (NLL), parameter count, and approximate training FLOPs to characterize accuracy-efficiency trade-offs. Our results show that attention-based models dominate MLPs in per-FLOP efficiency even at small scale, while increasing depth or context without sufficient optimization can degrade performance. We further examine rotary positional embeddings (RoPE), finding that architectural techniques successful in large language models do not necessarily transfer to small-model regimes.

Related papers

• ShishuLM: Lightweight Language Model with Hybrid Decoder-MLP Architecture and Paired Weight Sharing [0.5565728870245015]
  We introduce an efficient language model architecture, referred to as ShishuLM, which reduces both the parameter count and Key-Value (KV) cache requirements.<n>Our results show that ShishuLM provides up to 25% reduction in memory requirements and up to 40% improvement in latency during both training and inference, compared to parent models.
  arXiv  Detail & Related papers  (2025-10-13T04:04:54Z)
• Latent Thought Models with Variational Bayes Inference-Time Computation [52.63299874322121]
  Latent Thought Models (LTMs) incorporate explicit latent thought vectors that follow an explicit prior model in latent space.<n>LTMs demonstrate superior sample and parameter efficiency compared to autoregressive models and discrete diffusion models.
  arXiv  Detail & Related papers  (2025-02-03T17:50:34Z)
• Efficient Language Modeling for Low-Resource Settings with Hybrid RNN-Transformer Architectures [8.442206285783463]
  Transformer-based language models have recently been at the forefront of active research in text generation.<n>These models' advances come at the price of prohibitive training costs, with parameter counts in the billions and compute requirements measured in petaflop/s-decades.<n>We investigate transformer-based architectures for improving model performance in a low-data regime by selectively replacing attention layers with feed-forward and quasi-recurrent neural network layers.
  arXiv  Detail & Related papers  (2025-02-02T01:05:09Z)
• Feature Alignment-Based Knowledge Distillation for Efficient Compression of Large Language Models [4.737806982257592]
  This study proposes a knowledge distillation algorithm based on large language models and feature alignment.<n>The proposed model performs very close to the state-of-the-art GPT-4 model in terms of evaluation indicators such as perplexity, BLEU, ROUGE, and CER.
  arXiv  Detail & Related papers  (2024-12-27T04:37:06Z)
• Fine-tuning large language models for domain adaptation: Exploration of training strategies, scaling, model merging and synergistic capabilities [4.389938747401259]
  This work explores the effects of fine-tuning strategies on Large Language Models (LLMs) in domains such as materials science and engineering. We find that the merging of multiple fine-tuned models can lead to the emergence of capabilities that surpass the individual contributions of the parent models.
  arXiv  Detail & Related papers  (2024-09-05T11:49:53Z)
• Revisiting SMoE Language Models by Evaluating Inefficiencies with Task Specific Expert Pruning [78.72226641279863]
  Sparse Mixture of Expert (SMoE) models have emerged as a scalable alternative to dense models in language modeling. Our research explores task-specific model pruning to inform decisions about designing SMoE architectures. We introduce an adaptive task-aware pruning technique UNCURL to reduce the number of experts per MoE layer in an offline manner post-training.
  arXiv  Detail & Related papers  (2024-09-02T22:35:03Z)
• Observational Scaling Laws and the Predictability of Language Model Performance [51.2336010244645]
  We propose an observational approach that bypasses model training and instead builds scaling laws from 100 publically available models. We show that several emergent phenomena follow a smooth, sigmoidal behavior and are predictable from small models. We show how to predict the impact of post-training interventions like Chain-of-Thought and Self-Consistency as language model capabilities continue to improve.
  arXiv  Detail & Related papers  (2024-05-17T17:49:44Z)
• The Role of Model Architecture and Scale in Predicting Molecular Properties: Insights from Fine-Tuning RoBERTa, BART, and LLaMA [0.0]
  This study introduces a systematic framework to compare the efficacy of Large Language Models (LLMs) for fine-tuning across various cheminformatics tasks. We assessed three well-known models-RoBERTa, BART, and LLaMA-on their ability to predict molecular properties. We found that LLaMA-based models generally offered the lowest validation loss, suggesting their superior adaptability across tasks and scales.
  arXiv  Detail & Related papers  (2024-05-02T02:20:12Z)
• QualEval: Qualitative Evaluation for Model Improvement [82.73561470966658]
  We propose QualEval, which augments quantitative scalar metrics with automated qualitative evaluation as a vehicle for model improvement. QualEval uses a powerful LLM reasoner and our novel flexible linear programming solver to generate human-readable insights. We demonstrate that leveraging its insights, for example, improves the absolute performance of the Llama 2 model by up to 15% points relative.
  arXiv  Detail & Related papers  (2023-11-06T00:21:44Z)
• The Languini Kitchen: Enabling Language Modelling Research at Different Scales of Compute [66.84421705029624]
  We introduce an experimental protocol that enables model comparisons based on equivalent compute, measured in accelerator hours. We pre-process an existing large, diverse, and high-quality dataset of books that surpasses existing academic benchmarks in quality, diversity, and document length. This work also provides two baseline models: a feed-forward model derived from the GPT-2 architecture and a recurrent model in the form of a novel LSTM with ten-fold throughput.
  arXiv  Detail & Related papers  (2023-09-20T10:31:17Z)
• RAVEN: In-Context Learning with Retrieval-Augmented Encoder-Decoder Language Models [57.12888828853409]
  RAVEN is a model that combines retrieval-augmented masked language modeling and prefix language modeling. Fusion-in-Context Learning enables the model to leverage more in-context examples without requiring additional training. Our work underscores the potential of retrieval-augmented encoder-decoder language models for in-context learning.
  arXiv  Detail & Related papers  (2023-08-15T17:59:18Z)

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.