arXiv subjects

Faezeh Dehghan Tarzjani

Publications and source records attributed to Faezeh Dehghan Tarzjani.

6 recordsLinked to original sources

arXiv2026

From Local Evidence to Safety Verdicts: Causal Tracing in Vision-Language Models

A vision-language model may need to combine an image with a prompt to recognize a safety risk that neither reveals alone. Where does this joint safety judgment become accessible inside the model? We introduce SSU-Bench, a dataset of matched safe and unsafe image-text combinations constructed using single-item prompt edits or image edits with annotated intended regions. Using three vision-language models, we transfer internal states between paired inputs and measure the resulting change in the safety verdict. Across models and both types of counterfactual, interventions at the changed input positions are effective in earlier decoder layers, while interventions at the final input token become effective later. Directions estimated from other examples produce similar late-layer effects. A linear readout of the final-token state also predicts the model's own verdict, including incorrect judgments, and cross-model comparisons reveal similarities in the patterns of counterfactual change. These findings identify a recurring transition in where interventions can influence a joint safety verdict and distinguish a readable model decision from a correct safety judgment.

cs.AI↗

arXiv2026

Evolving Inspectable O-RAN Slicing xApps with LLMs

Open RAN (O-RAN) slicing xApps must adapt resource allocations to changing channel conditions and traffic demands while meeting service-level agreements (SLAs). Deep reinforcement learning can produce adaptive policies, but their allocation rules remain encoded in neural-network parameters. Our goal is to retain this adaptability while making the controller's decision logic directly inspectable and editable by operators. We use a large language model (LLM) to evolve slicing controllers as compact Python programs whose decision logic remains readable and editable after optimization. The LLM proposes and revises candidates offline, while a calibrated simulator scores them, and the selected decision module runs unchanged in the O-RAN control path. On the NSF POWDER 5G testbed, the evolved controller releases resources from a guaranteed slice whose throughput target becomes unattainable under a sustained channel fade, improving best-effort throughput from 158.2 to 228.6 Mbps, a 44.5% gain over the best static allocation. Since the controllers are readable source code, their behavior can be predicted from their equations, defects can be diagnosed by reading the code, and calibration errors can be corrected with one-line edits, reducing SLA misses from 79.9% to 2.2% in one case and more than doubling fitness in another. In a four-slice trace-driven simulation calibrated to the same testbed, evolutionary search achieves higher average evaluation scores than independent prompting at a matched proposal budget, with mean normalized gains on held-out traces of 16.3% for prompting alone, 32.1% for evolution from scratch, and 51.0% for evolution from a starting program.

eess.SY↗

arXiv2026

Spillover Effects under Network Interference When Neighbours' Treatment Effects Are Heterogeneous

Optimising budget-constrained network interventions requires evaluating not just who is connected, but predicting how strongly individual recipients will propagate the treatment's benefits. Existing models predict spillover from neighbours' treatments and attributes. We argue that spillover also depends on how strongly each neighbour responded to its own treatment. We prove that no model in which neighbours' responsiveness enters separately from their treatments can capture this interaction, and we introduce \textsc{SpilloverNet}, a graph neural network designed to preserve neighbour-level response dynamics. Because a neighbour's response is unobserved, a natural approach is to estimate it from covariates and plug it in. However, we prove that any predictor relying solely on standard network data faces an irreducible error floor set by unobserved personal responsiveness. Empirically, at low heterogeneity the plug-in's correlation with the true response looks acceptable while its spillover error is already 24.5\% against an oracle 13.8\%; as heterogeneity grows its error climbs to 51.9\%, worse than using no responsiveness estimate at all. A per-unit estimate from a direct-response measurement, collected in a small pilot before spillover arrives, escapes this bound and cuts regret against oracle-$τ$ targeting by up to 14 percentage points in a budgeted targeting problem. On two real social graphs, \textsc{SpilloverNet} reaches 7.7--8.4\% error, outperforming standard GNNs as well as specialised causal-representation baselines.

cs.CE↗

arXiv2025

Learning Wireless Interference Patterns: Decoupled GNN for Throughput Prediction in Heterogeneous Multi-Hop p-CSMA Networks

The p-persistent CSMA protocol is central to random-access MAC analysis, but predicting saturation throughput in heterogeneous multi-hop wireless networks remains a hard problem. Simplified models that assume a single, shared interference domain can underestimate throughput by 48-62% in sparse topologies. Exact Markov-chain analyses are accurate but scale exponentially in computation time, making them impractical for large networks. These computational barriers motivate structural machine learning approaches like GNNs for scalable throughput prediction in general network topologies. Yet off-the-shelf GNNs struggle here: a standard GCN yields 63.94% normalized mean absolute error (NMAE) on heterogeneous networks because symmetric normalization conflates a node's direct interference with higher-order, cascading effects that pertain to how interference propagates over the network graph. Building on these insights, we propose the Decoupled Graph Convolutional Network (D-GCN), a novel architecture that explicitly separates processing of a node's own transmission probability from neighbor interference effects. D-GCN replaces mean aggregation with learnable attention, yielding interpretable, per-neighbor contribution weights while capturing complex multihop interference patterns. D-GCN attains 3.3% NMAE, outperforms strong baselines, remains tractable even when exact analytical methods become computationally infeasible, and enables gradient-based network optimization that achieves within 1% of theoretical optima.

cs.LG↗

arXiv2025

An Efficient Lightweight Blockchain for Decentralized IoT

The Internet of Things (IoT) is applied in various fields, and the number of physical devices connected to the IoT is increasingly growing. There are significant challenges to the IoT's growth and development, mainly due to the centralized nature and large-scale IoT networks. The emphasis on the decentralization of IoT's architecture can overcome challenges to IoT's capabilities. A promising decentralized platform for IoT is blockchain. Owing to IoT devices' limited resources, traditional consensus algorithms such as PoW and PoS in the blockchain are computationally expensive. Therefore, the PoA consensus algorithm is proposed in the blockchain consensus network for IoT. The PoA selects the validator as Turn-based selection (TBS) that needs optimization and faces system reliability, energy consumption, latency, and low scalability. We propose an efficient, lightweight blockchain for decentralizing IoT architecture by using virtualization and clustering to increase productivity and scalability to address these issues. We also introduce a novel PoA based on the Weight-Based-Selection (WBS) method for validators to validate transactions and add them to the blockchain. By simulation, we evaluated the performance of our proposed WBS method as opposed to TBS. The results show reduced energy consumption, and response time, and increased throughput.

cs.CR↗

arXiv2025

Computing the Saturation Throughput for Heterogeneous p-CSMA in a General Wireless Network

A well-known expression for the saturation throughput of heterogeneous transmitting nodes in a wireless network using p-CSMA, derived from Renewal Theory, implicitly assumes that all transmitting nodes are in range of, and therefore conflicting with, each other. This expression, as well as simple modifications of it, does not correctly capture the saturation throughput values when an arbitrary topology is specified for the conflict graph between transmitting links. For example, we show numerically that calculations based on renewal theory can underestimate throughput by 48-62% for large packet sizes when the conflict graph is represented by a star topology. This is problematic because real-world wireless networks, such as wireless IoT mesh networks, are often deployed over a large area, resulting in non-complete conflict graphs. To address this gap, we present a computational approach based on a novel Markov chain formulation that yields the exact saturation throughput for each node in the general network case for any given set of access probabilities, as well as a more compact expression for the special case where the packet length is twice the slot length. Using our approach, we show how the transmit probabilities could be optimized to maximize weighted utility functions of the saturation throughput values. This would allow a wireless system designer to set transmit probabilities to achieve desired throughput trade-offs in any given deployment.

cs.NI↗