arXiv subjects

Ayesha Mohsin

Publications and source records attributed to Ayesha Mohsin.

5 recordsLinked to original sources

arXiv2026

Expertise-Based Developer Assignment for Long-Term Software Components in Open-Source Projects

Open-source software development through GitHub has enabled countless software projects to be developed by developers from across the world. Assigning the right task to the right developer enables teams to work efficiently. Assigning a task which is not representative of a developer's expertise results in slower completion time and less maintainable code. Whilst much work has been done to automate this assignment for specific jobs, there is little work addressing the assignment of developers to expertise-specific long-term components of a new project. This paper produces a web application that models developers' expertise based on their previous Git commits and automatically assigns them to an optimal task within a new project. Testing showed that the system's speed varies depending on the back-end model used but server-side data caching improved worst-case speeds by a factor of 9.86. 72.4% of tasks had their target developer listed in the top 10, out of a possible 47.

cs.SE↗

arXiv2026

What If We Allocate Test-Time Compute Adaptively?

Test-time compute scaling allocates inference computation uniformly, uses fixed sampling strategies, and applies verification only for reranking. In contrast, we propose a verifier-guided adaptive framework treating reasoning as iterative trajectory generation and selection. For each problem, the agent runs multiple inference iterations. In each iteration, it optionally produces a high-level plan, selects a set of reasoning tools and a compute strategy together with an exploration parameter, and then generates a candidate reasoning trajectory. A process reward model (PRM) serves as a unified control signal: within each iteration, step-level PRM scores are aggregated to guide pruning and expansion during generation, and across iterations, aggregated trajectory rewards are used to select the final response. Across datasets, our dynamic, PRM-guided approach consistently outperforms direct test-time scaling, yielding large gains on MATH-500 and several-fold improvements on harder benchmarks such as AIME24 and AMO-Bench. We characterize efficiency using theoretical FLOPs and a compute intensity metric penalizing wasted generation and tool overhead, demonstrating that verification-guided allocation concentrates computation on high-utility reasoning paths.

cs.CL↗

arXiv2026

Epistemic Uncertainty for Test-Time Discovery

Automated scientific discovery using large language models relies on identifying genuinely novel solutions. Standard reinforcement learning penalizes high-variance mutations, which leads the policy to prioritize familiar patterns. As a result, the maximum reward plateaus even as the average reward increases. Overcoming this limitation requires a signal that distinguishes unexplored regions from intrinsically difficult problems. This necessitates measuring disagreement across independently adapted weight hypotheses rather than relying on a single network's confidence. UG-TTT addresses this challenge by maintaining a small ensemble of low-rank adapters over a frozen base model. The per-token disagreement, quantified as the mutual information between ensemble predictions and weight hypotheses, isolates epistemic uncertainty and identifies positions where insufficient coverage leads to adapter divergence rather than intrinsic problem difficulty. This measure is incorporated as an exploration bonus into the policy gradient, directing the policy toward positions where persistent adapter disagreement signals low training coverage, the same frontier where genuine discovery is possible. A nuclear norm regularizer ensures the adapters remain distinct from one another, thereby preserving the exploration signal throughout training. Across four scientific discovery benchmarks, UG-TTT increases the maximum reward on three tasks, maintains substantially higher solution diversity, and an ablation study confirms that the regularizer is essential for sustaining this behavior.

cs.LG↗

arXiv2026

Towards Green Connectivity: An AI-Driven Mesh Architecture for Sustainable and Scalable Wireless Networks

Traditional macro-cell and micro-cell infrastructures suffer from severe inefficiencies, with current macro-cell networks operating at less than 5 percent energy efficiency, leading to nearly 95 percent of RF power wasted in covering vacant areas. The problem becomes particularly acute in high-density scenarios such as the Hajj, where approximately 7,000 temporary diesel-powered towers are deployed each year, consuming 56 million liters of fuel and emitting around 148,000 tons of CO2, yet still experiencing failure rates of nearly 40 percent at peak demand. To overcome these limitations, we propose an AI-driven mesh architecture based on three integrated enablers: (i) proximity-based deployment of low-power nodes within 250 to 300 meters of users, yielding a 38 dB link-budget gain and up to 6000 times efficiency improvement; (ii) spatial frequency reuse, which partitions cells into multiple non-interfering zones and achieves nearly 20 times capacity gain; and (iii) predictive network intelligence leveraging LSTMs to forecast traffic 5 seconds ahead, enabling smarter allocation and reducing congestion by about 60 percent. System-level evaluations combining propagation modeling and validated link-budget analysis demonstrate that this architecture delivers up to an 84 times improvement in useful energy delivery, reduces deployment costs by nearly 74 percent, and eliminates diesel dependence through solar-powered operations, thereby enabling sustainable, green connectivity for both rural and ultra-dense urban environments.

cs.NI↗

arXiv2026

Uncertainty-Weighted Experience Replay for Continual MIMO Channel Prediction

In dynamic wireless environments, accurate channel state information (CSI) prediction remains challenging due to non-stationary fading, mobility. This paper proposes an Uncertainty-Weighted Experience Replay (UW-ER) framework that integrates model uncertainty into the replay sampling process to improve robustness in online CSI prediction. A lightweight LSTM architecture with Monte-Carlo dropout is employed to estimate predictive variance, which is then used to adaptively weight the reconstruction loss for each training sample. The proposed method is evaluated on a UMi-Dense MIMO channel dataset generated using a stochastic fading model consistent with 3GPP standards. Results show that UW-ER achieves stable generalization, with validation NMSE centered near 0 dB and a strong correlation (r = 0.93) between predicted uncertainty and reconstruction error, indicating well-calibrated confidence estimates. Ablation studies demonstrate that the LARS-based replay policy achieves competitive performance with smaller memory budgets compared to conventional reservoir replay. Overall, the UW-ER approach improves continual channel learning stability without increasing computational complexity, offering a scalable solution for future 6G adaptive communication systems.

eess.SP↗