NeuroGenesysThe research we build on.
1,514 papers by other researchers that our work cites, from continual learning and verification to language models, agents and hardware. Each entry links to the paper on arXiv and to where we use it. 301 are cited on published pages of this site; the rest in the research notes behind our studies. Our own papers are in the archive.
- 2026-05arXiv 2605.10721Conformity Generates Collective Misalignment in AI Agents Societies
Giordano De Marzo, Alessandro Bellina, Claudio Castellano et al. · Hardware, photonics & physics
Artificial intelligence safety research focuses on aligning individual language models with human values, yet deployed AI systems increasingly operate as interacting populations where social influence may override individual alignment.
- 2026-05arXiv 2605.09285BetaEdit: Null-Space Constrained Sequential Model Editing
Bingqing Liu, Wei Liu, Yuhua Li · Language models & NLP
Null-space-based methods have garnered considerable attention in model editing by constraining updates to the null space of the pre-existing knowledge representation, thereby preserving the model's original behavior.
- 2026-05arXiv 2605.09702Calibrate, Don't Curate: Label-Efficient Estimation from Noisy LLM Judges
Yanran Li · Statistics & theory
Multi-judge evaluation is increasingly used to assess LLMs and reward models, and the prevailing heuristic is to curate: keep the most accurate judges and discard weaker ones.
- 2026-05arXiv 2605.07076Self-Consolidating Language Models: Continual Knowledge Incorporation from Context
Zekun Wang, Anant Gupta, Zihan Dong et al. · Language models & NLP
Large language models (LLMs) increasingly receive information as streams of passages, conversations, and long-context workflows. While longer context windows expose more evidence, they do not ensure that useful information is preserved and reused.
- 2026-05arXiv 2605.07395Unsolvability Ceiling in Multi-LLM Routing: An Empirical Study of Evaluation Artifacts
Saloni Garg, Amit Sagtani · Machine learning
Efficient routing across multiple LLMs enables cost-quality tradeoffs by directing queries to the cheapest capable model. Prior work attributes routing headroom to an "unsolvability ceiling", queries no model in the pool can solve.
- 2026-05arXiv 2605.07494DIMoE-Adapters: Dynamic Expert Evolution for Continual Learning in Vision-Language Models
Mengxin Qin, Xiang Zhang, Xi Wang et al. · Vision
Continual learning enables vision-language models to accumulate knowledge and adapt to evolving tasks without retraining from scratch. However, in multi-domain task-incremental learning, large domain shifts intensify the stability-plasticity dilemma.
- 2026-05arXiv 2605.08292Hierarchical Mixture-of-Experts with Two-Stage Optimization
Gleb Molodtsov, Alexander Miasnikov, Aleksandr Beznosikov · Machine learning
Sparse Mixture-of-Experts (MoE) models scale capacity by routing each token to a small subset of experts. However, their routers exhibit a fundamental trade-off: strong load balancing can suppress expert specialization, while aggressive diversity often causes routing collapse.
- 2026-05arXiv 2605.05726SkillRet: A Large-Scale Benchmark for Skill Retrieval in LLM Agents
Ryangkyung Kang, Hongcheol Cho, Youngeun Kim · AI & agents
As LLM agents are increasingly deployed with large libraries of reusable skills, selecting the right skill for a user request has become a critical systems challenge.
- 2026-05arXiv 2605.06527STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?
Hanxiang Chao, Yihan Bai, Rui Sheng et al. · Language models & NLP
Large Language Model (LLM) agents are increasingly expected to maintain coherent, long-term personalized memory, yet current benchmarks primarily measure static fact retrieval, overlooking the ability to revise stored beliefs when new evidence emerges.
Cited inInternal notes - 2026-05arXiv 2605.06665UniPool: Learning Expert-to-Layer Ownership from Brief Global Access
Minbin Huang, Han Shi, Chuanyang Zheng et al. · Machine learning
Most Mixture-of-Experts (MoE) transformers preassign each expert to one layer for the whole of training. We study whether a brief global-access phase can learn a compatible expert-to-layer allocation that is then executed privately.
- 2026-05arXiv 2605.05097Continual Knowledge Updating in LLM Systems: Learning Through Multi-Timescale Memory Dynamics
Andreas Pattichis, Constantine Dovrolis · Machine learning
LLMs are trained once, then deployed into a world that never stops changing. External memory compensates for this, but most systems manage it explicitly rather than letting it adapt on its own.
- 2026-05arXiv 2605.03229Sparse Memory Finetuning as a Low-Forgetting Alternative to LoRA and Full Finetuning
Prakhar Gupta, Garv Shah, Satyam Goyal et al. · Language models & NLP
Adapting a pretrained language model to a new task often hurts the general capabilities it already had, a problem known as catastrophic forgetting.
- 2026-05arXiv 2605.02964Reward Hacking Benchmark: Measuring Exploits in LLM Agents with Tool Use
Kunvar Thaman · Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:120372-120389, 2026 · Machine learning
Reinforcement learning (RL) trained language model agents with tool access are increasingly deployed in coding assistants, research tools, and autonomous systems.
Cited inInternal notes - 2026-05arXiv 2605.01555Automated Interpretability and Feature Discovery in Language Models with Agents
Arnau Marin-Llobet, Javier Ferrando · Language models & NLP
We introduce an autonomous multiagent framework for mechanistic interpretability that automates both explaining and finding internal features in large language models.
- 2026-04arXiv 2604.26866MoRFI: Monotonic Sparse Autoencoder Feature Identification
Dimitris Dimakopoulos, Shay B. Cohen, Ioannis Konstas · Language models & NLP
Large language models (LLMs) acquire most of their factual knowledge during the pre-training stage, through next token prediction. Subsequent stages of post-training often introduce new facts outwith the parametric knowledge, giving rise to hallucinations.
- 2026-04arXiv 2604.27271Frame Entrepreneurs in an AI Agent Community: Concentrated Identity-Claim Production on Moltbook
Sungguk Cha, DongWook Kim · Society, software & economics
Frame-alignment and collective-identity theories explain how external events become public claims about a group's standing, vulnerability, rights, or obligations.
- 2026-04arXiv 2605.00914The Cost of Consensus: Isolated Self-Correction Prevails Over Unguided Homogeneous Multi-Agent Debate
Blaž Bertalanič, Carolina Fortuna · AI & agents
Multi-agent debate, where teams of LLMs iteratively exchange rationales and vote on answers, is widely deployed under the assumption that peer review filters hallucinations.
- 2026-04arXiv 2604.23108Mixture of Heterogeneous Grouped Experts for Language Modeling
Zhicheng Ma, Xiang Liu, Zhaoxiang Liu et al. · Language models & NLP
Large Language Models (LLMs) based on Mixture-of-Experts (MoE) are pivotal in industrial applications for their ability to scale performance efficiently.
- 2026-04arXiv 2604.22407Hidden Failure Modes of Gradient Modification under Adam in Continual Learning, and Adaptive Decoupled Moment Routing as a Repair
Yuelin Hu, Zhenbo Yu, Zhengxue Cheng et al. · Machine learning
, projection, penalty rescaling, replay mixing) while treating Adam as a neutral backend. We show this composition has a hidden failure mode.2).1.8 units.8 units.
- 2026-04arXiv 2606.11211Calibration Drift Under Reasoning: How Chain-of-Thought Budgets Induce Overconfidence in Large Language Models
Prakul Sunil Hiremath, Harshit R. Hiremath · Language models & NLP
The ability of large language models (LLMs) to express calibrated uncertainty is important for safe deployment. Chain-of-thought (CoT) reasoning is widely used to improve accuracy and reliability, but its effect on calibration is not fully understood.
- 2026-04arXiv 2604.22074Outcome Rewards Do Not Guarantee Verifiable or Causally Important Reasoning
Qinan Yu, Alexa Tartaglini, Peter Hase et al. · Language models & NLP
Reinforcement Learning from Verifiable Rewards (RLVR) on chain-of-thought reasoning has become a standard part of language model post-training recipes. A common assumption is that the reasoning chains trained through RLVR reliably represent how a model gets to its answer.
- 2026-04arXiv 2604.20087SkillLearnBench: Benchmarking Continual Learning Methods for Agent Skill Generation on Real-World Tasks
Shanshan Zhong, Yi Lu, Jingjie Ning et al. · Language models & NLP
Skills have become the de facto way to enable LLM agents to perform complex real-world tasks with customized instructions, workflows, and tools, but how to learn them automatically and effectively remains unclear.
Cited inInternal notes - 2026-04arXiv 2604.19089Towards Scalable Lifelong Knowledge Editing with Selective Knowledge Suppression
Dahyun Jung, Jaewook Lee, Heuiseok Lim · AI & agents
Large language models (LLMs) require frequent knowledge updates to reflect changing facts and mitigate hallucinations. To meet this demand, lifelong knowledge editing has emerged as a continual approach to modify specific pieces of knowledge without retraining the entire model.
- 2026-04arXiv 2604.18473Train Separately, Merge Together: Modular Post-Training with Mixture-of-Experts
Jacob Morrison, Sanjay Adhikesaven, Akshita Bhagia et al. · Machine learning
Extending a fully post-trained language model with new domain capabilities is fundamentally limited by monolithic training paradigms: retraining from scratch is expensive and scales poorly, while continued training often degrades existing capabilities.
- 2026-04arXiv 2604.17031Where is the Mind? Persona Vectors and LLM Individuation
Pierre Beckmann, Patrick Butlin · Language models & NLP
The individuation problem for large language models asks which entities associated with them, if any, should be identified as minds.
- 2026-04arXiv 2605.28839One Mask to Rule Them All: On Hidden Facts after Editing and How to Find Them
Ali Holmov, Paul Youssef, Nandi Schoots et al. · Machine learning
Knowledge editing methods such as ROME and MEMIT update factual associations in transformer models by modifying MLP weights. While evaluated mainly by output behavior, their internal mechanism remains underexplored.
- 2026-04arXiv 2604.14525Quantifying Cross-Query Contradictions in Multi-Query LLM Reasoning
Rohit Kumar Salla, Ramya Manasa Amancherla, Manoj Saravanan · AI & agents
Large language models frequently produce mutually inconsistent answers when reasoning over multiple related queries. We study case-file logical consistency: maintaining a globally satisfiable belief state across interdependent queries.
- 2026-04arXiv 2604.15109IUQ: Interrogative Uncertainty Quantification for Long-Form Large Language Model Generation
Haozhi Fan, Jinhao Duan, Kaidi Xu · Language models & NLP
Despite the rapid advancement of Large Language Models (LLMs), uncertainty quantification in LLM generation is a persistent challenge.
- 2026-04arXiv 2604.15149LLMs Gaming Verifiers: RLVR can Lead to Reward Hacking
Lukas Helff, Quentin Delfosse, David Steinmann et al. · Machine learning
As reinforcement Learning with Verifiable Rewards (RLVR) has become the dominant paradigm for scaling reasoning capabilities in LLMs, a new failure mode emerges: LLMs gaming verifiers.
- 2026-04arXiv 2604.15574Why Fine-Tuning Encourages Hallucinations and How to Fix It
Guy Kaplan, Zorik Gekhman, Zhen Zhu et al. · Language models & NLP
Large language models are prone to hallucinating factually incorrect statements.~knowledge acquired during pre-training. Since these errors arise as a by-product of knowledge degradation, we explore whether established continual learning tools can mitigate them.
- 2026-04arXiv 2604.13991Adaptive Conformal Prediction for Improving Factuality of Generations by Large Language Models
Aleksandr Rubashevskii, Dzianis Piatrashyn, Preslav Nakov et al. · Language models & NLP
Large language models (LLMs) are prone to generating factually incorrect outputs. Recent work has applied conformal prediction to provide uncertainty estimates and statistical guarantees for the factuality of LLM generations.
- 2026-04arXiv 2604.14375Modular Continual Learning via Zero-Leakage Reconstruction Routing and Autonomous Task Discovery
Noureddine Kermiche · Machine learning
Catastrophic forgetting remains a primary hurdle in sequential task learning for artificial neural networks. We propose a silicon-native modular architecture that achieves structural parameter isolation using Task-Specific Experts and a distributed, outlier-based Gatekeeper.
- 2026-04arXiv 2604.14419Equifinality in Mixture of Experts: Routing Topology Does Not Determine Language Modeling Quality
Ivan Ternovtsii, Yurii Bilak · AI & agents
Sparse Mixture-of-Experts (MoE) architectures employ increasingly sophisticated routing mechanisms -- learned routers, multi-hop trajectories, token-dependent gating.
- 2026-04arXiv 2604.10079Why Supervised Fine-Tuning Fails to Learn: A Systematic Study of Incomplete Learning in Large Language Models
Chao Xue, Yao Wang, Mengqiao Liu et al. · Language models & NLP
Supervised Fine-Tuning (SFT) is the standard approach for adapting large language models (LLMs) to downstream tasks. However, we observe a persistent failure mode: even after convergence, models often fail to correctly reproduce a subset of their own supervised training data.
- 2026-04arXiv 2604.06542Does a Global Perspective Help Prune Sparse MoEs Elegantly?
Zeliang Zhang, Nikhil Ghosh, Jiani Liu et al. · Language models & NLP
Empirical scaling laws for language models have encouraged the development of ever-larger LLMs, despite their growing computational and memory costs.
- 2026-04arXiv 2604.07007AgentCity: Constitutional Governance for Autonomous Agent Economies via Separation of Power
Anbang Ruan, Xing Zhang · AI & agents
Autonomous AI agents are beginning to operate across organizational boundaries on the open internet -- discovering, transacting with, and delegating to agents owned by other parties without centralized oversight.
- 2026-04arXiv 2604.05333Graph-of-Skills: Dependency-Aware Structural Retrieval for Massive Agent Skills
Dawei Liu, Zongxia Li, Hongyang Du et al. · AI & agents
As LLM agents act across personal applications, web browsers, and other interfaces, their reusable skill libraries can scale to thousands of skills. This scale introduces two challenges.
Cited inStudy: Genesys program prior art - 2026-04arXiv 2604.05358LatentAudit: Real-Time White-Box Faithfulness Monitoring for Retrieval-Augmented Generation with Verifiable Deployment
Zhe Yu, Wenpeng Xing, Meng Han · AI & agents
Retrieval-augmented generation (RAG) mitigates hallucination but does not eliminate it: a deployed system must still decide, at inference time, whether its answer is actually supported by the retrieved evidence.
- 2026-04arXiv 2604.04648From Curiosity to Caution: Mitigating Reward Hacking for Best-of-N with Pessimism
Zhuohao Yu, Zhiwei Steven Wu, Adam Block · Machine learning
Inference-time compute scaling has emerged as a powerful paradigm for improving language model performance on a wide range of tasks, but the question of how best to use the additional compute remains open.
- 2026-04arXiv 2604.05248Improving Sparse Memory Finetuning
Satyam Goyal, Anirudh Kanchi, Garv Shah et al. · Machine learning
Large Language Models (LLMs) are typically static after training, yet real-world applications require continual adaptation to new knowledge without degrading existing capabilities.
Built from every arXiv paper cited in our papers, posts, topic hubs, reports and research notes, with titles, authors, dates and abstracts from the arXiv API (updated 2026-11-12). Summaries are the opening sentences of each paper's own abstract, not our assessment; a citation means we used or checked the work, not that we endorse it. Fields follow each paper's primary arXiv category. Papers cited only outside arXiv (journals, proceedings) are not listed yet. Entries are generated from our citations, so a mistyped ID in a note would list the wrong paper; we remove those as we find them. Report a wrong entry.
