CS Paper Radar

Weekly top 20

2026-W33 Aug 10 – Aug 16, 2026

Ranked from 3,675 papers submitted on 7 days. Each research area gets slots in proportion to its share of this week's papers (at least one), filled by score; near-duplicates are removed. Why

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

VGGT-Align: Bridging Local Reconstruction and Global Consistency for Long-Sequence 3D Reconstruction

cs.CV cs.AI ·2026-08-15 ·Wei Zhang, Yihang Wu, Songhua Li, Qi Wang
top venuecode/datahighly cited
Abstract · 10 pages, 6 figures, 6 tables. ACM Multimedia 2026 (MM '26). Code: https://github.com/WZ-CS/VGGT-Align

Maintaining global geometric consistency is a central challenge in long-sequence 3D reconstruction, with scale drift being the most critical failure mode. In chunk-based inference pipelines, the scale degree of freedom in sequential Sim(3) alignment is left unconstrained, causing estimation errors to compound multiplicatively and distort global trajectories and point cloud geometry. We present a scale-consistency enhancement framework built on a key insight: in structured environments such as driving scenes, geometric quantities arising from environmental regularity remain inherently invariant across temporal segments, and discrepancies in their per-chunk measurements directly expose inter-chunk scale drift. We propose Scene Geometric Invariant Anchoring (SGIA), which extracts dominant geometric invariants from each chunk's predicted point cloud via coarse-to-fine robust estimation and exploits their cross-chunk consistency to establish scale constraints independent of point cloud registration, explicitly degenerating 7-DoF Sim(3) alignment into 6-DoF rigid-body transformation and severing chain-wise scale error propagation at its source. We further introduce a lightweight test-time adaptation strategy that fine-tunes only normalization-layer parameters via multi-objective self-supervision, progressively improving intra-chunk predictions along the sequence. Both modules are plug-and-play and require no offline retraining. Experiments on multiple long-sequence benchmarks demonstrate state-of-the-art performance, reducing absolute trajectory error by up to 32% with significant gains in trajectory stability and reconstruction quality. Code: https://github.com/WZ-CS/VGGT-Align

0.99
2

LookBack: Where and How to Score LVLM Responses via Visual Reference Usage

cs.CV cs.AIcs.CL ·2026-08-12 ·Beomsik Cho, Jinhyeong Kim, Dongseok Lee, Jaehyung Kim
top venuecode/data
Abstract · Findings of EMNLP 2026. 24 pages, 12 figures. Code: https://github.com/bscho333/LookBack

Large Vision-Language Models (LVLMs) integrate visual perception with language generation, enabling responses that span image understanding and complex reasoning. However, LVLMs do not just inherit the text-level hallucinations; they also hallucinate against the image, producing fluent responses ungrounded in what they see. This makes LVLM response scoring inherently harder, and our diagnostics show that existing confidence-based metrics adopted from LLMs are insufficient for LVLMs. Specifically, removing the input image barely changes confidence-based selection, suggesting that output-space confidence primarily captures textual plausibility rather than agreement with the image. To address this gap, we propose LookBack, a training-free LVLM response scoring method that augments token likelihood with visual lookback score, a lightweight measure of how strongly each response token refers to image tokens. Across four benchmarks and three models, LookBack consistently improves Best-of-$N$ selection over existing baselines with negligible additional overhead.

0.96
3

SpeedTuning: Speeding Up Policy Execution with Lightweight Reinforcement Learning

cs.RO cs.AI ·2026-08-10 ·David D. Yuan, Tony Z. Zhao, Kaylee Burns, Chelsea Finn
top venuecode/datahighly cited
Abstract · 10 pages, 12 figures. This arXiv version includes an appendix with qualitative simulation rollouts and additional ablations. Published at IC…

While learned robotic policies hold promise for advancing generalizable manipulation, their practical deployment is often hindered by suboptimal execution speeds. Imitation learning policies are inherently limited by hardware constraints and the speed of the operator during data collection. In addition, there are no established methods for accelerating policies learned via imitation, and the empirical relationship between execution speed and task success remains underexplored. To address these issues, we introduce SpeedTuning, a reinforcement learning framework specifically designed to enhance the speed of manipulation policies. SpeedTuning learns to predict the optimal execution speed for actions, thereby complementing a base policy without necessitating additional data collection. We provide empirical evidence that SpeedTuning achieves substantial improvements in execution speed, exceeding 2.4x speed-up, while preserving an adequate success rate compared to both the original task policy and straightforward speed-up methods such as linear interpolation at a fixed speed. We evaluate our approach across a diverse set of dynamic and precise tasks, including pouring, throwing, and picking, demonstrating its effectiveness and robustness in enhancing real-world robotic manipulation. Videos and code are available at https://daivdyuan.github.io/speed-tuning/

0.95
4

One Frozen Simulator Is Not Enough: Simulator Collapse in Multi-Agent RL

cs.CL cs.AIcs.LG ·2026-08-12 ·Simon Yu, Nicholas Tomlin, Marwa Abdulhai, Ximing Lu, Derek Chong +5 more
top venuecode/datahighly cited
Abstract · 42 pages, 29 figures

Multi-agent reinforcement learning for human-AI interaction typically relies on a single large language model to simulate user behavior. We show that this approach systematically fails to generalize, and trace the failure to simulator collapse: because the simulator LLM is mode-collapsed, an LLM policy trained against it overfits to narrow strategies that exploit the simulator's dominant mode, and such a policy transfers poorly to unseen simulators and real users. We formalize this collapse theoretically and propose two complementary solutions, one at inference time and one at training time. The inference-time solution, Verbalized Sampling, broadens the simulator's behavior by sampling from a verbalized response distribution, reducing mode collapse. The training-time solution, Co-Training, jointly optimizes the policy against a population of trainable simulators, preventing it from overfitting to any single simulator's mode. We validate both solutions on three multi-turn benchmarks: Persuasion for Good, $τ^2$-bench, and CooperBench. Verbalized Sampling improves held-out success by up to 9% over single-simulator RL, and Co-Training pushes gains further to 14%; the human study shows similar gain on real users. Both solutions preserve the policy diversity that collapses under single-simulator RL. To support further work in this direction, we release SCOPE, an open-source framework for Population Co-Training multi-agent RL. More broadly, our results suggest that the diversity of the training environment, not only the policy, is critical to the generalization of multi-turn RL to real-world deployment.

0.95
5

A Survey of Large Models in Sports

cs.CL cs.CV ·2026-08-14 ·Yichen Xu, Jianzhe Ma, Chuhan Wang, Zhonghao Cao, Liangyu Chen +2 more
top venuesurveycode/data
Abstract · 36 pages, 4 figures, 6 tables. Accepted to Findings of ACL 2026

Sports have witnessed growing global enthusiasm in recent years, serving as a vital force for physical health, cultural exchange, social connection, and economic growth. The rapid advancement of large models, particularly (multimodal) large language models (M)LLMs, has demonstrated transformative potential to reshape sports understanding, analysis, and interaction across diverse domains. This paper presents a comprehensive survey of large models in sports, including (i) an overview of tasks and applications across different participant groups; (ii) a detailed analysis of sports-related datasets and benchmarks; and (iii) a critical discussion of current challenges and future directions. Our goal is to establish a foundation for advancing research and practical development of large-model-driven sports intelligence. An open-source GitHub repository is maintained at: https://github.com/Road2Redemption/Awesome_Large_Models_In_Sports1.

0.95
6

Nonlinear Model Predictive Control of a Robotic Soft Esophagus

cs.RO ·2026-08-10 ·Dipankar Bhattacharya, Ryman Hashem, Leo K. Cheng, Weiliang Xu
top venuecode/datahighly cited
Abstract · Accepted manuscript. 12 pages. Published in IEEE Transactions on Industrial Electronics. Project page: https://bhattner143.github.io/rosev2-…

Strictures caused by esophageal cancer can narrow down the esophageal lumen, leading to dysphagia. Palliation of dysphagia has driven the development of a Robotic Soft Esophagus (RoSE), which provides a novel in vitro platform for esophageal stent testing and food viscosity studies. In RoSE, peristaltic wave generation and control were done in an open-loop manner since the conduit lacked visibility and embedded sensing capability. Hence, in this work, RoSE version 2.0 (RoSEv2.0) is designed with embedded Time Of Flight (TOF) and pressure sensors to measure conduit displacement and air pressure, respectively, for modeling and control. Model Predictive Control (MPC) of RoSEv2.0 is implemented to govern the peristalsis and air pressure profile autonomously. The implemented MPC used Sparse Identification Nonlinear Dynamics with Control (SINDYC) models to estimate the future states of ROSEv2.0. The dynamic models are discovered from the TOF and pressure sensor data. Peristalsis waves of speed 20 mm/s, wavelength 75 mm, and amplitudes 5, 7.5, and 10 mm were successfully generated by the MPC. Additionally, RoSEv2.0 with the MPC was employed to perform stent migration testing with various food bolus consistencies. The major contribution claimed in this paper is the application of SINDYC-based MPC to solve the closed-loop control problem of RoSE for achieving desired peristaltic waves.

0.94
7

Evo-Harness: Context-to-Harness Skill Compilation for Self-Evolving Agents

cs.AI cs.CL ·2026-08-15 ·Tianxin Wei, Zhan Shi, Minhua Lin, Bing He, Zewen Liu +12 more
top venuecode/datahighly cited
Abstract · EMNLP 2026 Main

Learning from experience is critical for developing capable, self-improving large language model (LLM) agents. Existing methods typically extract knowledge from accumulated trajectories via reflection, memory, rules, or skills. However, agents in realistic environments continuously encounter novel tasks, often offering only a one-shot opportunity to improve. These executions yield rich but highly noisy contexts, entangling broadly useful lessons with task-specific artifacts. Critically, prior works rarely validate their effectiveness on complex real-world tasks or isolate the underlying drivers of improvement. To address these gaps, we formulate online harness learning, where a frozen agent improves by continually updating a structured harness across sequential tasks. This formulation enables a systematic study of key self-improvement factors through our proposed Evo-Harness. At its core, context-to-harness skill compilation distills noisy, single-shot executions into reusable skill harnesses for cross-domain and topic-level adaptation. To demonstrate the efficacy of one-shot skill compilation, we evaluate across five realistic benchmarks (TerminalBench2, SWE-bench, CL-Bench, -bench, WebArena-Infinity). Our extensive analysis demonstrates the effectiveness of Evo-Harness and provides a principled understanding of how LLM agents can effectively learn on the fly. Our code is available at https://github.com/A-EVO-Lab/a-evolve/tree/release/evo-harness.

0.94
8

VisEditBench: Can Vision-Language Models Edit Visualization Code from Multimodal Feedback?

cs.CL ·2026-08-11 ·Mizanur Rahman, Arshia Azimlu, Shadikur Rahman, Md Tahmid Rahman Laskar, Amran Bhuiyan +2 more
top venuebenchmarkcode/datahighly cited
Abstract

Vision-language models (VLMs) have shown strong capabilities in generating visualization code from textual or visual specifications. However, real-world visualization authoring is inherently iterative: users frequently revise existing visualizations to repair flawed charts or adapt them to desired styles. Existing benchmarks primarily evaluate generation from scratch, leaving visualization code editing from multimodal feedback largely unexplored. We introduce VisEditBench, a benchmark of 1,395 human-annotated visualization code-editing tasks grounded in realistic visualization workflows and failure cases. VisEditBench covers two practical settings: feedback-guided repair, where models revise visualization code using buggy or marked charts together with textual feedback, and reference-guided restyling, where models modify code to match a target chart image. Evaluating 20 state-of-the-art VLMs reveals that visualization code editing remains challenging: Claude-4.6-Sonnet achieves the best overall pass rate of 74.46%, while most open-source models remain below 50%. Performance is particularly weak on visually grounded style adaptation, where Claude-4.6-Sonnet achieves only 55.71%. To establish a strong baseline, we further propose VisEditAgent, a render-grounded editing framework that iteratively generates, executes, validates, and refines candidate edits. Built on GPT-4o, VisEditAgent improves overall pass rate from 55.75% to 67.99%, demonstrating the importance of render-grounded feedback for faithful visualization editing. We will release VisEditBench at https://github.com/vis-nlp/VisEditBench.

0.93
9

Gaze Target Estimation Anywhere with Concepts

cs.CV cs.AI ·2026-08-11 ·Xu Cao, Houze Yang, Vipin Gunda, Zhongyi Zhou, Tianyu Xu +3 more
top venuebenchmarkcode/datahighly cited
Abstract · CVPR 2026 Code and Benchmark are aviliable at https://github.com/IrohXu/GazeAnywhere and https://huggingface.co/datasets/IrohXu/Gaze-Co-Benc…

Estimating human gaze targets from images in-the-wild is an important and formidable task. Existing approaches primarily employ brittle, multi-stage pipelines that require explicit inputs, like head bounding boxes and human pose, in order to identify the subject of gaze analysis. As a result, detection errors can cascade and lead to failure. Moreover, these prior works lack the flexibility of specifying the gaze analysis task via natural language prompting, an approach which has been shown to have significant benefits in convenience and scalability for other image analysis tasks. To overcome these limitations, we introduce the Promptable Gaze Target Estimation (PGE) task, a new end-to-end, concept-driven paradigm for gaze analysis. PGE conditions gaze prediction on flexible user text or visual prompts (e.g., "the boy in the red shirt" or "person in point [0.52, 0.48]") to identify a specific subject for gaze analysis. This approach integrates subject localization with gaze estimation, and eliminates the rigid dependency on intermediate analysis stages. We develop a scalable data engine to generate Gaze-Co (Gaze Estimation with Concepts), a dataset and benchmark of 120K high-quality, prompt-annotated image pairs. We also propose GazeAnywhere, the first model designed for PGE. GazeAnywhere uses a transformer-based detector to fuse features from frozen encoders and simultaneously solves subject localization, in/out-of-frame presence, and gaze target heatmap estimation. GazeAnywhere achieves state-of-the-art performance on multiple PGE benchmarks, setting a strong baseline for this new problem even on a difficult out-of-domain, real-world clinical dataset. GazeAnywhere is open-sourced in github.com/IrohXu/GazeAnywhere.

0.92
10

PatientAct: Theory-Grounded Mental Health Client Simulation

cs.CL cs.AIcs.HC ·2026-08-13 ·Sahand Sabour, TszYam NG, Yaqian Chen, Guanqun Bi, Jialu Zhao +1 more
top venuecode/datahighly cited
Abstract · EMNLP 2026 Findings

LLM-based simulated clients are increasingly used to train novice counselors, evaluate LLM therapists, and generate synthetic data. However, current simulators produce overly cooperative clients that disclose too readily, accept therapeutic reframes without resistance, and resolve core issues within a single session. We trace these issues to profiles that lack causal depth and behavioral mechanisms that treat all content as equally accessible. We present PatientAct, a framework for client simulation grounded in established clinical theories. Our profiles integrate the 5Ps clinical case formulation, providing causal depth without tying the design to any single therapeutic modality. During simulation, profiles include a dynamic memory layer in which items carry trust thresholds (e.g., symptoms are available early, whereas formative memories require a sustained therapeutic alliance). At each turn, the client's emotional reaction and behavior are modeled before generating a response. If the therapist approaches gated content, PatientAct expresses resistance in terms of quantity, content, and style rather than defaulting to cooperation or a single resistance pattern. We evaluate our framework on 40 clinical situations and demonstrate that it generates diverse profiles with high clinical plausibility. Moreover, PatientAct significantly outperforms the baselines, yielding substantial gains in resistance quality and behavioral realism. Our code and data are publicly available via github.com/Sahandfer/PatientHub.

0.92
11

Sycophants in the Courtroom: Are LLMs Fragile to Juridical Authority and Evolving Legal Standards?

cs.CY cs.AIcs.CL ·2026-08-10 ·Lorenzo Molfetta, Alessio Cocchieri, Luca Ragazzi, Ilaria Bartolini, Marco Patella +1 more
top venuebenchmarkhighly cited
Abstract · Please cite the definitive, peer-reviewed version of this article published in the Proceedings of the 64th Annual Meeting of the Association…

In medicine, claims remain valid when supported by empirical evidence grounded in stable biological reality. In law, by contrast, truth is contingent, defined by jurisdiction, temporal validity, and the hierarchy of authoritative sources. The recent success of large language models (LLMs) on medical licensing examinations has encouraged an expectation of comparable legal competence. This analogy, however, obscures a critical distinction between domains. Unlike in medicine, legal performance often depends less on inference than on determining when external authority is applicable, valid, and non-contradictory. We introduce a comparative diagnostic framework evaluating legal reasoning against medical baselines along four axes (knowledge recall, grounding, confidence, and robustness), uncovering a sharp domain asymmetry when applied to a new benchmark that encodes temporal validity and normative relationships. While medical LLMs reliably benefit from verified sources, legal LLMs struggle to assess when retrieved citations are useful or misleading, exhibiting overconfidence in perturbed contexts and sensitivity to superficial formatting cues. Increased model scale amplifies this tendency, revealing that stronger instruction following can coincide with weaker resistance to authoritative perturbations. These findings show that LLMs treat law as unstructured text rather than binding precedent, while revealing a tendency to over-trust authoritative but false information when external references conflict with a model's internal knowledge.

0.90
12

CForce: Boosting Parallel Decoding for dLLMs via Consistency Forcing

cs.LG cs.AIcs.CL ·2026-08-14 ·Yuji Ren, Chenkai Xu, Zhuocheng Gong, Jianguo Li, Zhijie Deng
top venuecode/data
Abstract

Diffusion large language models (dLLMs) accelerate language generation by predicting multiple masks in a single forward pass. However, existing dLLMs can suffer from unreliable predictions in early denoising stages under aggressive parallelism strategies, leading to errors that can propagate to later stages. To tackle this issue, we present Consistency Forcing (CForce) for dLLMs, a distillation method to force the mask predictions of early stages to align with those of later stages. CForce trains the model on pre-collected self-rollout trajectories, thereby improving training-inference alignment. We introduce Confidence Adaptive KL Divergence as a distillation objective to conjoin the merits of forward and reverse KL. We further provide a theoretical analysis for the consistency objective to explain why CForce can approximately minimize the prediction error of early stages. Critically, the same formulation applies to both mask-to-token decoding and edit-capable decoding; in the edit-capable case, later token-to-token refinements provide additional supervision for earlier masked-state predictions. Experiments on non-edit and edit-capable LLaDA models show improved speed-quality trade-offs, especially under high-parallelism decoding budgets. Code is available at: https://github.com/inclusionAI/dFactory.

0.88
13

PACE-Bench: Benchmarking Physics Adaptation via Code Evolution in Dynamic Environments

cs.AI ·2026-08-14 ·Yuhao Zhan, Bingxiang He, Zecong Tang, Chaojun Xiao
top venuebenchmarkcode/data
Abstract

Self-evolving agents improve future behavior from interaction experience, yet existing evaluations typically optimize under fixed execution conditions and do not test recovery after those conditions change. To address this gap, we introduce PACE-Bench (Physics Adaptation via Code Evolution), a simulator-grounded benchmark of 144 source-to-target adaptation pairs across six physics domains. Each pair links a source environment to a mutated target environment with the same goal and interface. A code-driven design that succeeds in the source fails in the target, where agents must iteratively adapt it into a working target design using diagnostic sandbox feedback within a limited attempt budget. We compare ten self-evolving methods from four paradigms. The benchmark remains far from saturated: Reflexion + Qwen3-14B succeeds on only 35.9\% of full-benchmark pairs, while GPT-5.5 solves 66.7\% of the Statics subset under the full budget. Together, these results show that simulator-grounded reflection is more reliable than unverified self-revision, while memory anchors agents to early designs and broad tree search explores without converging. Even revealing exact physical changes does not raise the performance ceiling, pointing to mechanism redesign rather than parameter inference as the central bottleneck. Data and code are available at https://github.com/thunlp/PACE-Bench.

0.88
14

DACRI: Decision-Aware Causal Intervention Ranking for Critical Supply Chains

cs.LG ·2026-08-11 ·Shiqi Huang, Jiani He, Dingyan Shang, Yihua Xu, Jize Li +2 more
top venuebenchmarkcode/data
Abstract · Accepted for presentation at the International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME 2026…

Detecting or attributing a supply-chain disruption is not the same as selecting the intervention that maximizes recoverable net value. We present CriticalSCM-Bench v1, a controlled synthetic benchmark with causal ground truth, paired factual/counterfactual rollouts, and an explicit net-value objective. Relative to a full-information train-selected static benchmark, LambdaMART improves median normalized net value by 5.7--16.2\%, with paired statistical support on the semiconductor and critical-material archetypes but not on digital infrastructure. On digital infrastructure, a domain-informed constant-buffer policy remains stronger, showing that greater model complexity is not uniformly justified. Across partial and delayed settings, LambdaMART retains 33--75\% of full-clamp value. Stress tests further show that intervention fidelity, timing, cost, and held-out disruptions can alter policy ordering. Critical materials show the weakest out-of-distribution retention. Separately, a guarded explanation study over 540 generations preserves every fixed intervention decision after deterministic validation and template fallback, although exact wording remains unstable. Within this controlled setting, the results identify regimes in which adaptive ranking adds value and those in which simpler structural policies remain preferable.

0.88
15

RealisticTritonBench: A Benchmark for Triton-Kernel Generation in Real-World AI Frameworks

cs.SE cs.AI ·2026-08-12 ·Jinjun Huang, Zhongzhen Wen, Tongtong Xu, Meng Yan, Xin Xia +1 more
top venuebenchmarkcode/data
Abstract · Accepted by ASE 2026

In modern AI frameworks, GPU kernels are key to overall system performance. Combining usability, portability, and near-handwritten CUDA performance, Triton is widely adopted for implementing GPU kernels. Recent advances show the potential of large language models (LLMs) to automatically generate Triton kernels, reducing the manual effort required from expert kernel developers. Several benchmarks evaluate LLM-generated Triton kernels. However, they suffer from three key limitations: (1) they restrict tasks to PyTorch-to-Triton translation, failing to reflect the diversity and complexity of real-world Triton tasks; (2) they evaluate only individual-kernel performance rather than end-to-end performance, the core criterion for real-world deployment in AI frameworks; and (3) they rely on manually written evaluation scripts for individual kernels, which may contain flaws that models can exploit to bypass correctness checks and obtain inflated scores. To address these limitations, we introduce RealisticTritonBench, the first benchmark to derive Triton kernel generation tasks from real-world pull requests in popular AI frameworks, enabling realistic, production-like evaluation. RealisticTritonBench systematically extracts PRs that modify Triton kernels from popular open-source AI frameworks and transforms them into generation tasks with concrete engineering contexts. Each task takes a natural language requirement as input and requires a corresponding Triton kernel implementation, with a complete and reproducible evaluation environment. Unlike prior benchmarks focused on isolated kernel performance, RealisticTritonBench integrates generated kernels into their original frameworks and evaluates them using end-to-end tests, enabling a more faithful assessment. We evaluate leading LLMs on RealisticTritonBench and find that they still struggle with real-world Triton kernel generation tasks.

0.87
16

DSAgentBench: Can Agents Automate End-to-End Data-Science Workflows in Real Computer Environments?

cs.AI cs.CL ·2026-08-11 ·Mizanur Rahman, Mohammed Saidul Islam, Ridwan Mahbub, Md Tahmid Rahman Laskar, Shafiq Joty +1 more
top venuebenchmarkcode/data
Abstract

Real-world data science involves long-horizon workflows that span data wrangling, exploration, modeling, visualization, and validation, and require coordinated use of tools such as notebooks, IDEs, terminals, browsers, and databases within real operating environments. Yet existing benchmarks lack real-computer interaction and do not evaluate whether agents can execute complete end-to-end data-science workflows in realistic computing environments, failing to capture the multi-stage, multi-tool nature of data-science practice. We introduce DSAgentBench, the first benchmark to evaluate whether agents can automate full data-science workflows inside real computer environments. DSAgentBench contains 275 diverse tasks covering the entire data-science life-cycle, reflecting the complexity and tool coordination required in practice. Each task requires grounding decisions in intermediate outputs and coordinated tool use, and includes a deterministic evaluator that verifies analytical correctness, visual outputs, and model performance rather than code-only execution. Our extensive experiments with 15 closed- and open-source models show that even the strongest agent, Claude-4.6-Sonnet, achieves only 56.70% task success, while all open-source agents remain below 1%, frequently failing at tool orchestration, OS grounding, and multi-step reasoning. These results reveal a substantial capability gap between current agentic systems and real data-science workflows, positioning DSAgentBench as a foundation for developing grounded, verifiable, autonomous data-science agents. We release DSAgentBench at https://github.com/vis-nlp/DSAgentBench.

0.84
17

Retrieval-Corrected Conformal Prediction for Time Series

cs.LG cs.AI ·2026-08-11 ·Sangjin Jin, Kangmin Kim, Junhyeong Lee, Yongjae Lee
top venuecode/data
Abstract · Accepted at the 35th ACM International Conference on Information and Knowledge Management (CIKM '26), Rome, Italy

Conformal prediction (CP) provides distribution-free prediction intervals for fixed forecasters, but its standard calibration procedure is often inefficient for time series data, where forecast errors are temporally dependent and change across time and operating conditions. Recent time series CP methods improve local calibration using recent, weighted, or localized residuals. Yet local calibration can remain indirect, since broad residual weighting or additional adaptation procedures may dilute the evidence most relevant to the current prediction. This motivates a simple retrieval and correction strategy that selects similar past residuals as local evidence and then corrects the coverage error left by retrieval. In this paper, we propose Retrieval--Corrected Conformal Prediction (RCCP), a retrieval-augmented calibration method for time series prediction intervals. RCCP builds an asymmetric interval from retrieved one-sided residuals and calibrates its normalized retrieval error with a scalar conformal correction. Thus, retrieval provides local residual evidence, while conformal correction determines the final scale needed for coverage. We provide a coverage-gap bound based on the stability of the normalized retrieval error distribution. Across standard benchmarks and backbone forecasters, RCCP attains the target coverage in every setting and achieves the lowest Winkler scores, with fewer severe misses. RCCP also achieves low calibration and inference overhead, showing that retrieval-corrected calibration is an effective and scalable approach to uncertainty quantification in time series forecasting. Code is available at https://github.com/jinsaaang/rccp.

0.84
18

Enabling Beyond-Visual-Line-of-Sight Drones Operation over Open RAN 5G Networks with Slicing

cs.NI ·2026-08-10 ·Pau Baguer, Esteban Municio, Gines Garcia-Aviles, Xavier Costa-Pérez
publishedcode/datahighly cited
Abstract · 7 pages, 4 figures. Published in IEEE Network

Among the foretold claims of the transition from 5G to 6G, Beyond-Visual-Line-of-Sight (BVLoS) drone operation has emerged as a prominent Internet-of-Robots enabler. However, safety concerns have been raised since BVLoS imposes strict requirements on performance and dependability on the technology, and requires robust regulatory frameworks. While current 5G technologies promise to meet the performance requirements in terms of throughput and latency, there is a lack of studies regarding how to achieve full reliability in practice. To address this challenge, the research community is actively working on open-source projects that allow for experimental validation in the field. Fortunately, new Open RAN (O-RAN) standards are paving the way for such approaches in an integrated, native manner. In this work, we deploy a state-of-the-art 5G O-RAN open-source BVLoS operational system, report current limitations, and address them via advanced capabilities natively available in O-RAN: Slicing. Our proposed deployment minimizes trajectory errors due to 5G link congestion and keeps latency well below the 3GPP limits defined for BVLoS operation. Finally, we discuss on the challenges ahead and the opportunities that 5G O-RAN-enabled networks may bring to BVLoS drone operation.

0.83
19

Optimal Pricing and Charging Strategy Design for Non-cooperative Battery Swapping Stations

cs.GT eess.SY ·2026-08-14 ·Huanyu Yan, Chenxi Sun, Huanxin Liao, Xiaoying Tang
top venuehighly cited
Abstract · Published in IEEE Transactions on Mobile Computing, vol. 23, no. 12, pp. 13573-13588, Dec. 2024

Battery swapping is a rapid way to recharge electric vehicles (EVs). As more and more entities are involved in building Battery Swapping Stations (BSSs), how non-cooperative BSSs maximize their profit in a competitive market needs further investigation. In this paper, we focus on a practical scenario where competitive BSSs are coordinated by the same aggregator. To study the optimal pricing and battery charging, we formulate a hierarchical game-theoretic model, where BSSs determine the swapping price in the day-ahead market in the first stage, and then determine the optimal battery charging strategy in the real-time market in the second stage. We rigorously prove the existence and uniqueness of the Subgame Perfect Nash Equilibrium (SPNE). In particular, the uniqueness property provides theoretical support that the strategy under equilibrium is optimal in the competitive environment. Based on the unique SPNE, we propose an optimal pricing and charging strategy for each BSS to maximize profit in the competitive market. A prediction error handling method is also proposed to deal with unexpected fluctuations in swapping demand. Our simulation with a 12-BSS system based on real-life data from Xi'an, China shows that our pricing and charging strategy increases the individual BSS profit by at least 18.1\%, while the optimal charging strategy naturally achieves peak shaving for the power grid.

0.75
20

Dynamic Deep Prompt Optimization for Defending Against Jailbreak Attacks on LLMs

cs.CR cs.AIcs.CL ·2026-08-10 ·Doniyorkhon Obidov, Honggang Yu, Xiaolong Guo, Kaichen Yang
top venuehighly cited
Abstract

Large Language Models (LLMs) demonstrate impressive capabilities across many applications but remain vulnerable to jailbreak attacks, which elicit harmful or unintended content. While model fine-tuning is an option for safety alignment, it is costly and prone to catastrophic forgetting. Prompt optimization has emerged as a promising alternative, yet existing prompt-based defenses typically rely on static modifications (e.g., fixed prefixes or suffixes) that cannot adapt to diverse and evolving attacks. We propose Dynamic Deep Prompt Optimization (DDPO), the first jailbreak defense based on deep prompt optimization. DDPO uses the target LLM's own intermediate layers as feature extractors to dynamically generate defensive embeddings via a lightweight multilayer perceptron. These tailored embeddings are then injected into a subsequent intermediate layer, enabling an input-dependent defense without modifying the LLM's weights. This design ensures high adaptability with minimal computational overhead. Experiments on a diverse set of models and attacks demonstrate that DDPO significantly outperforms static prompt optimization methods, particularly on weakly aligned models and when handling semantically ambiguous benign prompts, successfully distinguishing them from genuinely harmful requests.

0.73