arXiv subjects

Steven Kolawole

Publications and source records attributed to Steven Kolawole.

13 recordsLinked to original sources

arXiv2026

What Survives When You Compress a Recursive Reasoner for the Edge?

Recursive reasoning models solve hard structured tasks with a few million parameters by iterating a latent state, but deploying them on edge hardware means compressing them -- and quantization noise compounds across recursive cycles rather than accumulating over output tokens, so single-pass intuitions fail. Here, we ask what survives. Across a full precision sweep and three tasks, aggressive compression preserves local prediction but destroys global reasoning: pruning, distillation, and linear attention drive puzzle-exact accuracy to zero while cell accuracy holds. Naive INT4 is different: its damage is architectural, striking MLP-mixing recursion but not attention on the same task, and per-channel calibration reverses it without retraining, where quantization-aware training does not. We also introduce carry-trajectory fidelity, the cosine similarity to the full-precision reasoning path, as a label-free signal that predicts both the damage and the recovery before any task evaluation, with a threshold that transfers across tasks. The result is a deployment recipe: flash-streamed embeddings remove a 99.4 MB bottleneck, one recursive cycle matches full-depth accuracy at 6x fewer FLOPs, and calibrated INT4 brings the backbone to 3.3 MB, inside a 4 MB budget.

cs.LG↗

arXiv2026

A Tilted Bowl Is Not a Slippery Slope: Compressing Looped Models

Looped models reason by applying the same block of weights many times, so compressing that block saves memory traffic on every loop. Compressed looped models, however, often collapse, and the collapse is usually blamed on rounding error that accumulates from loop to loop. In this work we test that account on more than 30 models from five families and find, to our surprise, that it holds only for loops that never settle. When a loop settles, a fixed rounding error does not accumulate. It moves the point where the loop settles, much as tilting a bowl moves where a ball comes to rest, and the answer is lost only when the shift is larger than the readout tolerates. This picture lets us predict which models fail from a single label-free measurement, and it tells us why failed models recover: their loops still settle, so a few final loops with 8-bit weights bring the answer back. Motivated by these findings, we build a controller that stops when the model's halting head fires and then finishes with 8-bit loops. On Sudoku-Extreme and Maze-Hard it beats fixed-depth inference by up to 15 points under a third of the weight traffic.

cs.LG↗

arXiv2026

EpiKV: Epiphany-Aware KV Cache Eviction Without the Attention Matrix

Reasoning models can generate chains of thought tens of thousands of tokens long, making the key--value (KV) cache that holds them a major bottleneck for inference throughput. Existing eviction policies for long reasoning traces typically rank cached tokens using attention weights, requiring access to the attention matrix and making them incompatible with fast inference kernels. In this work we study the limits of such policies under tight cache budgets. Surprisingly, we find that under the strongest of them the generations that finish are wrong about as often as without eviction; most of the accuracy loss comes from generations that enter loops and run until the length limit, and retaining more tokens according to a fixed importance score exacerbates this behavior. What stops the looping is keeping the tokens the model's recent queries point to, and the forward pass the model already runs reveals them without the attention matrix. Motivated by this observation, we introduce epiphany-aware KV cache eviction EpiKV, which combines hidden-state shifts with the model's recent query--key relevance to rank cached tokens without materializing the attention matrix. On multiple benchmarks, EpiKV matches or outperforms the strongest attention-based eviction baselines while running directly in vLLM with unmodified attention kernels.

cs.LG↗

arXiv2026

PrivacyBench: Privacy Isn't Free in Hybrid Privacy-Preserving Vision Systems

Privacy preserving machine learning deployments in sensitive deep learning applications; from medical imaging to autonomous systems; increasingly require combining multiple techniques. Yet, practitioners lack systematic guidance to assess the synergistic and non-additive interactions of these hybrid configurations, relying instead on isolated technique analysis that misses critical system level interactions. We introduce PrivacyBench, a benchmarking framework that reveals striking failures in privacy technique combinations with severe deployment implications. Through systematic evaluation across ResNet18 and ViT models on medical datasets, we uncover that FL + DP combinations exhibit severe convergence failure, with accuracy dropping from 98% to 13% while compute costs and energy consumption substantially increase. In contrast, FL + SMPC maintains near-baseline performance with modest overhead. Our framework provides the first systematic platform for evaluating privacy-utility-cost trade-offs through automated YAML configuration, resource monitoring, and reproducible experimental protocols. PrivacyBench enables practitioners to identify problematic technique interactions before deployment, moving privacy-preserving computer vision from ad-hoc evaluation toward principled systems design. These findings demonstrate that privacy techniques cannot be composed arbitrarily and provide critical guidance for robust deployment in resource-constrained environments.

cs.CR↗

arXiv2026

Everybody Prune Now: Structured Pruning of LLMs with only Forward Passes

Structured pruning is a promising approach to create smaller, faster large language models. However, existing methods typically rely on computing the gradient via backward passes, which can inflate memory requirements and compute costs. In this work we introduce Bonsai, a gradient-free structured pruning method that eliminates the need for backpropagation, significantly reducing memory requirements and compute costs while achieving state-of-the-art pruning performance. Bonsai uses forward-pass-only perturbative pruning to enable efficient compression of large models on a broader range of hardware configurations. Unlike existing structured pruning approaches, Bonsai not only achieves better compression with fewer resources but also produces models that are twice as fast as those generated by semi-structured pruning. As a concrete demonstration, we use Bonsai to prune 7B and 8B models to 50% sparsity on a single A6000 GPU -- a task challenging for backprop-based methods in memory-constrained settings, as they require 2-3x the memory. Our results show that removing backprop as a requirement not only enables pruning larger models on constrained hardware but can also lead to state-of-the-art efficiency and performance.

cs.LG↗

arXiv2025

Semantic Agreement Enables Efficient Open-Ended LLM Cascades

Cascade systems route computational requests to smaller models when possible and defer to larger models only when necessary, offering a promising approach to balance cost and quality in LLM deployment. However, they face a fundamental challenge in open-ended text generation: determining output reliability when generation quality lies on a continuous spectrum, often with multiple valid responses. To address this, we propose semantic agreement -- meaning-level consensus between ensemble outputs -- as a training-free signal for reliable deferral. We show that when diverse model outputs agree semantically, their consensus is a stronger reliability signal than token-level confidence. Evaluated from 500M to 70B-parameter models, we find that semantic cascades match or surpass target-model quality at 40% of the cost and reduce latency by up to 60%. Our method requires no model internals, works across black-box APIs, and remains robust to model updates, making it a practical baseline for real-world LLM deployment.

cs.CL↗

arXiv2025

PARALLELPROMPT: Extracting Parallelism from Large Language Model Queries

LLM serving systems typically treat user prompts as monolithic inputs, optimizing inference through decoding tricks or inter-query batching. However, many real-world prompts contain latent semantic parallelism--decomposable structures where subtasks can be executed independently to reduce latency while preserving meaning. We introduce PARALLELPROMPT, the first benchmark for measuring intra-query parallelism in natural user prompts. Our dataset comprises over 37,000 real-world prompts from public LLM chat logs, each annotated with a structured schema capturing task templates, shared context, and iteration inputs. These schemas are extracted using LLM-assisted prompting with rule-based multilingual validation. To evaluate the benefits of decomposition, we provide an execution suite that benchmarks serial vs. parallel strategies, measuring latency, structural adherence, and semantic fidelity. Our results show that intra-query parallelism can be successfully parsed in over 75% of curated datasets, unlocking up to 5x speedups on tasks like translation, comprehension, and comparative analysis, with minimal quality degradation. By releasing this benchmark, curation pipeline, and evaluation suite, we provide the first standardized testbed for studying structure-aware execution in LLM serving pipelines.

cs.LG↗

arXiv2025

From Scarcity to Efficiency: Investigating the Effects of Data Augmentation on African Machine Translation

The linguistic diversity across the African continent presents different challenges and opportunities for machine translation. This study explores the effects of data augmentation techniques in improving translation systems in low-resource African languages. We focus on two data augmentation techniques: sentence concatenation with back translation and switch-out, applying them across six African languages. Our experiments show significant improvements in machine translation performance, with a minimum increase of 25\% in BLEU score across all six languages. We provide a comprehensive analysis and highlight the potential of these techniques to improve machine translation systems for low-resource languages, contributing to the development of more robust translation systems for under-resourced languages.

cs.CL↗

arXiv2025

Agreement-Based Cascading for Efficient Inference

Adaptive inference schemes reduce the cost of machine learning inference by assigning smaller models to easier examples, attempting to avoid invocation of larger models when possible. In this work we explore a simple, effective adaptive inference technique we term Agreement-Based Cascading (ABC). ABC builds a cascade of models of increasing size/complexity, and uses agreement between ensembles of models at each level of the cascade as a basis for data-dependent routing. Although ensemble execution introduces additional expense, we show that these costs can be easily offset in practice due to large expected differences in model sizes, parallel inference execution capabilities, and accuracy benefits of ensembling. We examine ABC theoretically and empirically in terms of these parameters, showing that the approach can reliably act as a drop-in replacement for existing models and surpass the best single model it aims to replace in terms of both efficiency and accuracy. Additionally, we explore the performance of ABC relative to existing cascading methods in three common scenarios: (1) edge-to-cloud inference, where ABC reduces communication costs by up to 14x; (2) cloud-based model serving, where it achieves a 3x reduction in rental costs; and (3) inference via model API services, where ABC achieves a 2-25x reduction in average price per token/request relative to state-of-the-art LLM cascades.

cs.LG↗

arXiv2024

What Happens When Small Is Made Smaller? Exploring the Impact of Compression on Small Data Pretrained Language Models

Compression techniques have been crucial in advancing machine learning by enabling efficient training and deployment of large-scale language models. However, these techniques have received limited attention in the context of low-resource language models, which are trained on even smaller amounts of data and under computational constraints, a scenario known as the "low-resource double-bind." This paper investigates the effectiveness of pruning, knowledge distillation, and quantization on an exclusively low-resourced, small-data language model, AfriBERTa. Through a battery of experiments, we assess the effects of compression on performance across several metrics beyond accuracy. Our study provides evidence that compression techniques significantly improve the efficiency and effectiveness of small-data language models, confirming that the prevailing beliefs regarding the effects of compression on large, heavily parameterized models hold true for less-parameterized, small-data models.

cs.CL↗

arXiv2023

Vision Transformers for Mobile Applications: A Short Survey

Vision Transformers (ViTs) have demonstrated state-of-the-art performance on many Computer Vision Tasks. Unfortunately, deploying these large-scale ViTs is resource-consuming and impossible for many mobile devices. While most in the community are building for larger and larger ViTs, we ask a completely opposite question: How small can a ViT be within the tradeoffs of accuracy and inference latency that make it suitable for mobile deployment? We look into a few ViTs specifically designed for mobile applications and observe that they modify the transformer's architecture or are built around the combination of CNN and transformer. Recent work has also attempted to create sparse ViT networks and proposed alternatives to the attention module. In this paper, we study these architectures, identify the challenges and analyze what really makes a vision transformer suitable for mobile applications. We aim to serve as a baseline for future research direction and hopefully lay the foundation to choose the exemplary vision transformer architecture for your application running on mobile devices.

cs.CV↗

arXiv2023

Adapting to the Low-Resource Double-Bind: Investigating Low-Compute Methods on Low-Resource African Languages

Many natural language processing (NLP) tasks make use of massively pre-trained language models, which are computationally expensive. However, access to high computational resources added to the issue of data scarcity of African languages constitutes a real barrier to research experiments on these languages. In this work, we explore the applicability of low-compute approaches such as language adapters in the context of this low-resource double-bind. We intend to answer the following question: do language adapters allow those who are doubly bound by data and compute to practically build useful models? Through fine-tuning experiments on African languages, we evaluate their effectiveness as cost-effective approaches to low-resource African NLP. Using solely free compute resources, our results show that language adapters achieve comparable performances to massive pre-trained language models which are heavy on computational resources. This opens the door to further experimentation and exploration on full-extent of language adapters capacities.

cs.CL↗

arXiv2021

Sign-to-Speech Model for Sign Language Understanding: A Case Study of Nigerian Sign Language

Through this paper, we seek to reduce the communication barrier between the hearing-impaired community and the larger society who are usually not familiar with sign language in the sub-Saharan region of Africa with the largest occurrences of hearing disability cases, while using Nigeria as a case study. The dataset is a pioneer dataset for the Nigerian Sign Language and was created in collaboration with relevant stakeholders. We pre-processed the data in readiness for two different object detection models and a classification model and employed diverse evaluation metrics to gauge model performance on sign-language to text conversion tasks. Finally, we convert the predicted sign texts to speech and deploy the best performing model in a lightweight application that works in real-time and achieves impressive results converting sign words/phrases to text and subsequently, into speech.

cs.CV↗