{"total":10,"items":[{"citing_arxiv_id":"2605.03279","ref_index":16,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"RFPrompt: Prompt-Based Expert Adaptation of the Large Wireless Model for Modulation Classification","primary_cat":"cs.LG","submitted_at":"2026-05-05T02:09:58+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"RFPrompt adapts the Large Wireless Model via deep prompt tokens to improve out-of-distribution robustness in modulation classification while training only a small number of parameters.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.27969","ref_index":51,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"From Mirage to Grounding: Towards Reliable Multimodal Circuit-to-Verilog Code Generation","primary_cat":"cs.SE","submitted_at":"2026-04-30T15:01:27+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":8.0,"formal_verification":"none","one_line_summary":"MLLMs exhibit a Mirage effect by bypassing circuit diagrams in favor of header semantics for Verilog generation; VeriGround with identifier anonymization and D-ORPO training reaches 46% Functional Pass@1 while refusing blank images at >92%.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.19488","ref_index":47,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"CoDA: Towards Effective Cross-domain Knowledge Transfer via CoT-guided Domain Adaptation","primary_cat":"cs.AI","submitted_at":"2026-04-21T14:10:37+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"CoDA aligns cross-domain latent reasoning representations in LLMs via CoT distillation and MMD to enable effective knowledge transfer without in-domain demonstrations.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.22823","ref_index":32,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"PivotMerge: Bridging Heterogeneous Multimodal Pre-training via Post-Alignment Model Merging","primary_cat":"cs.CV","submitted_at":"2026-04-18T09:38:03+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"PivotMerge merges heterogeneous multimodal pre-trained models via shared-space decomposition to filter conflicts and layer-wise weights based on alignment contributions, outperforming baselines on multimodal benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.10029","ref_index":34,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Self-Distilled Reinforcement Learning for Co-Evolving Agentic Recommender Systems","primary_cat":"cs.IR","submitted_at":"2026-04-11T04:52:38+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"CoARS enables co-evolving recommender and user agents by using interaction-derived rewards and self-distilled credit assignment to internalize multi-turn feedback into model parameters, outperforming prior agentic baselines.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.09462","ref_index":44,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Adaptor: Advancing Assistive Teleoperation with Few-Shot Learning and Cross-Operator Generalization","primary_cat":"cs.RO","submitted_at":"2026-04-10T16:19:43+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Adaptor uses few-shot learning with trajectory perturbation and vision-language conditioning to achieve robust cross-operator intent recognition and higher success rates in assistive teleoperation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.06696","ref_index":21,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"AgentGate: A Lightweight Structured Routing Engine for the Internet of Agents","primary_cat":"cs.AI","submitted_at":"2026-04-08T05:22:16+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"AgentGate decomposes routing into action decision and structural grounding stages, allowing small 3B-7B models to dispatch queries competitively on a curated benchmark after targeted fine-tuning.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.09688","ref_index":79,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Immunizing 3D Gaussian Generative Models Against Unauthorized Fine-Tuning via Attribute-Space Traps","primary_cat":"cs.CV","submitted_at":"2026-04-06T09:30:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"GaussLock embeds traps targeting position, scale, rotation, opacity, and color in 3D Gaussian models to degrade unauthorized fine-tunes while preserving authorized performance.","context_count":1,"top_context_role":"method","top_context_polarity":"use_method","context_text":"quaternion, andc∈R 3 is the color of each primitive. In our security context, the pre-trained weightsθserve as the primary intellectual property. We consider a threat model where an adversary attempts to perform unauthorized fine-tuning onθ to extract proprietary capabilities. To comprehensively demon- strate defensive robustness, we evaluate against both prevalent Low-Rank Adaptation (LoRA) [79] and full-parameter fine- tuning. Under the primary LoRA assumption, the adversary introduces low-rank update matrices∆θ=AB ⊤, where A∈R d×r andB∈R k×r are trainable matrices with rank r≪min(d, k). The objective of GaussLock is to embed a dormant defense mechanism directly into this parameter space to disrupt these unauthorized adaptations. C. Multi-Attribute Defensive Traps"},{"citing_arxiv_id":"2604.03066","ref_index":158,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"Redefining End-of-Life: Intelligent Automation for Electronics Remanufacturing Systems","primary_cat":"eess.SY","submitted_at":"2026-04-03T14:40:24+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":2.0,"formal_verification":"none","one_line_summary":"A literature review of intelligent automation approaches using robotics, AI, and control for disassembly, inspection, sorting, and reprocessing of end-of-life electronics.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"sufficient relevant knowledge and only requires clearer task specification. On the other hand, parameter-efficient fine- tuning [157] adapts a pretrained model to a specific domain that may not have been covered during pretraining. Instead of full fine-tuning of large models, which can be expensive in terms of memory, computation, and storage, methods such as LoRA [158] are widely used for LLM adaptation. The key idea is to preserve the pretrained backbone while learning a lightweight task-specific module using data from downstream tasks, making adaptation more practical. On top of LoRA- based adaptation, human preferences can also be incorporated through reinforcement learning from human feedback (RLHF) [159], in which the model is further tuned using feedback"},{"citing_arxiv_id":"2604.14179","ref_index":37,"ref_count":1,"confidence":0.55,"is_internal_anchor":false,"paper_title":"An Underexplored Frontier: Large Language Models for Rare Disease Patient Education and Communication -- A scoping review","primary_cat":"cs.CL","submitted_at":"2026-03-30T17:14:48+00:00","verdict":"ACCEPT","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A scoping review of 12 studies finds LLM applications for rare disease patient education remain early-stage, dominated by general models like ChatGPT focused on curated question-answering with limited real-world or patient-centered evaluation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}