CS Paper Radar

Weekly top 20 · cs.ET

2026-W47 Nov 16 – Nov 22, 2026

Ranked from 1,465 papers submitted on 2 days — week in progress, ranks can still change.

Score = weighted mean ofCitations 20%Upvotes 10%Code stars 10%Venue 15%Code/data 10%Novelty 15%Momentum 15%Substance 5%how it works
1

Noise Characterization and Robust Signal Detection in Yeast Pheromone Molecular Communication

cs.IT cs.ET ·2026-11-17 ·Nikolaos Ntetsikas, Styliana Kyriakoudi, Antonis Kirmizis, Ioannis Krikidis, Ian F. Akyildiz +1 more
publishedunusual
Abstract

A critical aspect of Molecular Communications (MC) is the implementation of signal detection policies amidst noise. To date, noise characterizations within the MC field have predominantly drawn from methodologies found in wireless communications literature. In this study, we diverge from existing MC research by utilizing a newly developed experimental platform that employs yeast, allowing us to consider more realistic noise characterizations based on the relevant signaling pathways. We propose suitable signal detection mechanisms tailored to this experimental setup, which focuses on yeast cell-to-cell communications. Our analysis identifies gene transcription as the primary source of noise, and we utilize a Markov birth-death process model with Poisson arrivals and departures to characterize it. The noisy expression of the FUS1 gene is best represented using a mixed Gaussian distribution model. This model serves as a foundation for evaluating the performance of Maximum Likelihood Detection mechanisms in terms of Bit Error Rate (BER) for both symbol-by-symbol and sequence transmission schemes. Error analysis indicates that appropriate adjustments to the signal threshold can reduce errors to as low as 10%, which is not negligible. In contrast, the detection of symbol sequences demonstrates enhanced error performance, achieving error rates as low as 0.4%, albeit at the cost of increased computational complexity.

0.64
2

SLED: Sensor SLicing from Edge to Datacenter

cs.AI cs.CEcs.DCcs.ET+1 ·2026-11-16 ·Alan Subedi, Avhishek Biswas, Apala Pramanik, Amlan Balabantaray, Mikhail Titov +3 more
top venuecode/data
Abstract · Accepted in IEEE Annual Congress on Artificial Intelligence of Things (IEEE AIoT) 2026. The SLED software and evaluation datasets are public…

Modern sensing systems generate large volumes of heterogeneous data streams from multiple sensor modalities, increasing reliance on datacenters for computation, artificial intelligence/superintelligence (AI/SI) inference, and AI/SI training. Each sensor stream may have different network and compute requirements depending on its type and data volume. However, current wireless systems typically allocate resources at the device level, leading to contention among co-located sensor streams, where bulk data transfers may starve latency-critical telemetry. To this end, this paper introduces Sensor Slicing (SL). SL provides deterministic, medium access control (MAC)-level isolation for individual sensors within a single user equipment (UE) in open radio access networks (O-RAN). Using agentic coding techniques, SL is implemented in two O-RAN-compliant packages, reducing development time by 400 times (2 h 40 min [OCUDU] vs. 6.5 weeks [OAI]), 228-fold fewer human prompts (4 vs. 913), and less than half the tokens (0.64B vs. 1.43B). An associated end-to-end framework, SLED, is implemented to establish per-sensor radio and network isolation across edge, campus, and remote high-performance computing (HPC) tiers. Over-the-air wireless experiments demonstrate that while conventional network slicing causes critical delays to increase more than 2,300-fold under high traffic load, SLED maintains stable latency. Scaling evaluations on the two RAN stacks show that throughput remains within 0.9% and 1.4% of expected allocations even as the number of sensor slices increases. The results demonstrate that SLED preserves latency-critical telemetry while effectively utilizing resources across the sensor-to-HPC pipeline. The SLED software and evaluation datasets are publicly available at https://github.com/UNL-CPN-Lab/SLED-Sensor-SLicing-from-Edge-to-Datacenter.

0.51
3

A Copper-Inducible Yeast Biosensor with Bioluminescent Readout: Mathematical Modeling, Experimental Validation, and Photodetector Coupling for Molecular Communication in IoBNT

cs.ET ·2026-11-17 ·Nikolaos Ntetsikas, Styliana Kyriakoudi, Ariel Klavaris, Antonis Kirmizis, Chrysostomos Chrysostomou +2 more
published
Abstract

Engineered living biosensors constitute promising model organisms for bringing Molecular Communications (MC) closer to practical applications. However, challenges associated with their development and characterization have resulted in the relevant literature being limited in the MC domain, . In this paper, we contribute in this direction by leveraging engineered yeast as a basis to detect heavy metals such as copper. The biosensor's development is supported by a validated mathematical characterization and analysis for possible integration in an MC system. Different from the existing literature, mathematical representations of the major biological processes within the yeast copper detoxification pathway are integrated into a comprehensive framework that predicts the time evolution of the CUP1 gene, which is treated as the system output. In the engineered circuit, CUP1 was replaced with GFP, therefore making GFP fluorescence the final system output. Experimental analysis including mRNA transcript quantification and fluorometric measurements, validate the GFP reporting model to a satisfactory degree. This establishes the proposed model as a useful tool for virtual experimentation and the developed testbed as a suitable platform for further engineering of yeast-based copper sensors. The theoretical analysis is further extended by incorporating a self-sustained bioluminescent reporter for yeast, replacing GFP and enabling optical output without external light excitation. The optical response is coupled to a photodetector model, rendering the system suitable for integration into mobile monitoring devices such as Smart Pills. The presented characterization of the yeast receiver as a copper biosensor, constitutes the first step towards the development and characterization of a complete end-to-end MC based system for gut theranostics after coupling with suitable engineered probiotic senders.

0.47
4

Living Context Graphs: Continual Semantic Learning for Enterprise Data Reasoning

cs.CL cs.AIcs.DBcs.ET+1 ·2026-11-16 ·Varun Agrawal, Ayush Srivastava, Eshan Gujar, Kaushal Kumar, Ayush Gupta
publishedcode/data
Abstract · 22 pages, 5 figures. Accepted at the NeurIPS 2026 Workshop on Continual Learning in the Era of Foundation Models and Embodied Agents. Data s…

Enterprise AI can now perform analytics, reporting, and data-driven workflows, yet the data foundations it relies on remain hard to operationalize: business meaning, relationships, metric definitions, and reporting conventions are absent from database schemas and scattered across undocumented systems and tribal knowledge. We introduce the Living Context Graph (LCG), a continually learned context layer that makes existing enterprise data operational for AI without requiring a copy or transformation of the underlying warehouse. LCG bootstraps base semantics and relationships from the data itself, builds a concept graph that indexes business context as queries are served, writes durable semantic learnings from verified outcomes, and withdraws or re-earns context when later outcomes contradict prior routes. We evaluate LCG through text-to-SQL, a core data-reasoning task, on a transfer-aware split of three BIRD databases comprising 499 questions: 353 form the learning history and 146 are held out and never used to update the context layer. Bootstrapping alone improves overall accuracy by +4.6 percentage points. Learning only from the history raises accuracy on the untouched held-out transfer set by +22.6 points (95% CI [+14.4, +30.8]), exceeding five-shot retrieval from the same history by +12.3 points. Incremental learning reaches the one-shot endpoint within observed evaluation noise on every database. Ablating the graph changes held-out accuracy by at most one question per database but increases per-query cost 2.3-3.3x and median latency 1.8-3.4x: continually learned semantics drive accuracy, while the graph makes that context efficient to serve.

0.43
5

MASQ: Measurement-Adaptive Sampling for Near-Clifford Quantum Error Correction Circuits

cs.ET quant-ph ·2026-11-17 ·Chaebin Jung, Sangmin Jeon, Kyeongwon Lee, Sechan Park, Woojoo Lee
published
Abstract · 7 pages, 3 figures, and 1 table. This paper has been accepted for publication at DATE 2027

Classical sampling of quantum error correction (QEC) circuits supports fault-tolerant protocol evaluation, but estimating rare logical failures requires many independent shots. Frame-active-state simulators isolate non-stabilizer information in a dense active-state vector, yet dense-state processing remains exponential in the active width. This paper presents MASQ, an exact measurement-adaptive sampler for near-Clifford QEC circuits based on selective materialization of non-Clifford rotations. At each measurement boundary, MASQ materializes only rotations in a measurement-relevant closure capturing direct and indirect non-commuting dependencies, deferring the rest to limit active-state expansion. Stabilizer-assisted elimination further removes rotations certified to be invisible to all future measurements. A circuit-dependent schedule is computed once and shared across shots, while adaptive caching of conditional Born probabilities and batched execution reduce repeated sampling work while preserving the measurement-record distribution. Across 19 QEC benchmarks, MASQ reduces peak active width by up to 40 active virtual qubits and achieves up to 8,098-fold CPU throughput speedup over Clifft on multi-round circuits.

0.43
6

3D-VASE: Volumetric-Aware Scalable and Efficient Multi-Layer Design Synthesis for 3D-Printed Microfluidics

cs.ET ·2026-11-17 ·Yushen Zhang, Jiashuo Chen, Tsun-Ming Tseng, Ulf Schlichtmann
published
Abstract · Accepted for publication in the 32nd Asia and South Pacific Design Automation Conference (ASP-DAC 2027)

Additive manufacturing (3D printing) enables complex multi-layer microfluidic devices, but the physical synthesis of passive-flow designs remains a key bottleneck. In such systems, designers must encode timing synchronization and target dosing directly into channel geometry, even though the available routing space across multiple 3D layers is often highly congested, which makes scaling to complex designs difficult. Existing approaches formulate placement, routing, and fluidic constraints as a single monolithic optimization, leading to combinatorial explosion and poor performance on complex designs. This paper presents a volumetric-aware two-stage synthesis framework that decouples global structure from routing detail. A Mixed-Integer Programming (MIP) model first computes a compact multi-layer placement, followed by a dynamic 3D routing engine that enforces timing and volumetric constraints during path construction. To further improve routability in dense layouts, we introduce a hierarchical conflict resolution strategy combining local rip-up-and-reroute with controlled global expansion. Experimental results show that, under timing and volumetric constraints, 3D-VASE consistently improves synthesis speed and, in most test cases, also improves layout quality. Notably, for difficult large-scale cases where prior methods often failed to return any solution within the time limit, 3D-VASE produces feasible high-quality layouts within practical runtimes. These advances enable fast, iterative design-space exploration for complex 3D-printed microfluidic systems.

0.36
7

Countering Anion Inhibition of Laccase Electron Transfer: Electrostatic Design, Charge-Transfer Dynamics on a Superconducting Quantum Processor, and Its Limits

cs.ET ·2026-11-17 ·Lucia Malickova
benchmarkunusual
Abstract · Main text 17 pages, supporting informations 13 pages

Hydroxide at neutral-to-alkaline pH and halides inhibit fungal laccases and limit their use in wastewater treatment. From GFN2-xTB cluster models of Trametes versicolor laccase (PDB 1GYC) we build a four-site Hubbard model of the His454-T3A-T3B-His452 junction of the trinuclear copper cluster and execute its charge-transfer dynamics on the Euro-Q-Exa quantum computer (IQM, 8 qubits, Trotterized circuits with Pauli twirling; 726 jobs). The charge-transfer frequency is measured to 3-5 meV, reproducibly over nine days and across changes of the device qubit map: hydroxide raises the charge-transfer energy of the crystal-geometry model by 92.0 +/- 2.2 meV (exact 87) and exceeds fluoride by 10.1 +/- 1.8 meV. A screen of charged probes in three dielectric models identifies Y116E; in the probe model the hardware resolves a reduction of the hydroxide penalty by 74 +/- 7 meV and of the halide penalty by 64 +/- 5 meV, acting on the site-energy gap rather than on the coupling. A real Glu116 (five conformers) with PROPKA protonation states and the coupled Asp456 gives a realistic reduction of 10-30 meV, with the same sign in 28 of 28 model combinations; the direction and ordering are the prediction, rate factors only an illustration. Scanning the on-site repulsion reverses the sign of the hydroxide effect, and a one-electron Hartree model reproduces the reversal: the measured dynamics is effectively single-particle. On a Fermi-Hubbard benchmark the hardware does not resolve electron correlation; it would need a contrast loss below 0.13 percent per CZ gate instead of the measured 0.34-0.54 percent.

0.32
8

Automated-Truck Penetration Shapes Human Drivers' Car-Following Margins Near Motorway Ramps: A Driving-Simulator Study

cs.ET ·2026-11-17 ·Minhao Chen, Zehao Jiang, Julius Secadiningrat, Elnaz Irannezhad
Abstract · 31 pages, 7 figures, 5 tables

Automated-vehicle dedicated lanes (AVDLs) are proposed as transitional infrastructure for mixed traffic, but their effect near ramps, and that of automated trucks, on the human drivers sharing the road has rarely been observed. In a mixed-design driving-simulator experiment, 63 drivers were assigned to one of four layouts (an AVDL extending beyond, coinciding with, or ending inside a 1.5-km ramp weaving section, or no AVDL) and drove at four automated-truck penetration levels from 20% to 80%. Linear mixed-effects models with participant random effects and multiplicity control were used. Higher penetration reduced the share of car-following time at gaps below 1.5 s from 28.5% to 18.1% (Holm-adjusted p = .006). This was explained by leader composition: drivers spent 8% of following time behind a truck at 20% penetration and 48% at 80%, kept longer gaps behind trucks, and did not change their gaps within leader type. AVDL layout had no detectable effect on safety margins, rear-end conflict rates, travel time or speed (differences bounded within plus or minus 5 s per section) and no significant effect on workload. Between-driver differences accounted for 52% to 77% of the variance in the main outcomes. Mixed-traffic models should therefore represent human car-following as leader-type-dependent and heterogeneous.

0.26
9

Residual-Guided Lightweight Reinforcement Learning for Risk-Aware Trajectory Selection with a Frozen World Model

cs.AI cs.ETcs.LGcs.RO+1 ·2026-11-17 ·Jiaqi Wu, Yongqi Dong, Yijie Zhong, Maple Shao, Yide Tao +4 more
Abstract · 6 pages, 2 figures

In automated driving, world models provide future scene predictions for evaluating candidate trajectories, but fixed selection rules may not capture scene-dependent trade-offs. We propose a residual-guided reinforcement learning (RL) framework for lightweight trajectory selection with a frozen world model and a fixed candidate bank. The framework uses proximal policy optimization (PPO) to fine-tune a joint policy over candidate selection and bounded trajectory residuals. A context-conditioned ranking network learns candidate preferences from feedback obtained by evaluating local perturbations of the selected trajectories. At inference, the selector combines learned preference and predicted risk ranks using fixed weights and returns the selected candidate trajectory without modification. Fine-tuning updates 16,106 parameters across the actor and critic, while the deployed selector retains only the 11,725-parameter ranking network, with the residual branch and critic used exclusively during training. In open-loop evaluation on the nuScenes validation set, the method reduces average Euclidean trajectory error (L2) by 3.99% relative to the RiskWorld model baseline without RL-based selection. Under matched budgets for feedback accesses and actor updates, residual exploration reduces L2 by 0.83% relative to the categorical PPO variant trained without residuals. The method also maintains lower L2 than RiskWorld when risk features are removed or shuffled.

0.26
10

Threat Landscape of Vibe Coding and the Road to Secure Autonomy

cs.CR cs.ET ·2026-11-16 ·Devki Nandan Jha, Shubh Anand, Karthick Keeranagere Krishnaiah, Roman Ahearn, Suhas Nale
Abstract · 35 pages, 6 figures, 5 tables, Submitted to ACM Computing Surveys

Vibe coding is reshaping software development, enabling faster implementation and allowing individuals with limited programming experience to describe requirements in natural language while AI agents write, test, and deploy the code. As these agents evolve into autonomous systems that execute commands, manage dependencies, access cloud resources, and interact with other agents, their security risks extend beyond incorrect code to actions with direct consequences for software systems. Although research on these risks is expanding, it remains fragmented across areas including prompt injection, insecure code generation, supply-chain attacks, and human factors, with limited consensus on terminology, threat assumptions, and evaluation. This paper consolidates emerging research through the lens of the vibe coding workflow. We establish a reference architecture, characterise six levels of agent autonomy, and develop a threat model centred on five trust boundaries spanning developers, AI models, tools, cloud infrastructure, and the wider software ecosystem. Building on this foundation, we propose a taxonomy of twenty-one threats across nine classes. For each threat, we review the available evidence, assess proposed defences, and identify limitations. Our analysis highlights that risks increase with agents' operational authority, while most defences rely on model judgement and address isolated threats. Finally, we identify eleven research challenges, including enforceable limits on agent authority and shared benchmarks and methodologies for assessing vibe coding security.

0.21