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.
- 2026-01arXiv 2601.21282WorldBench: Benchmarking Physical Understanding of World Models by Isolating Physics Concepts
Rishi Upadhyay, Howard Zhang, Jim Solomon et al. · Vision
Recent advances in generative foundational models, often termed "world models," have propelled interest in applying them to critical tasks like robotic planning and autonomous system training.
- 2026-01arXiv 2601.21375TeachBench: A Syllabus-Grounded Framework for Evaluating Teaching Ability in Large Language Models
Zheng Li, Siyao Song, Jingyuan Ma et al. · AI & agents
Large language models (LLMs) show promise as teaching assistants, yet their teaching capability remains insufficiently evaluated. Existing benchmarks mainly focus on problem-solving or problem-level guidance, leaving knowledge-centered teaching underexplored.
- 2026-01arXiv 2601.21577Collaborative Parameter Learning: Mitigating Forgetting via Parameter-Level Gradient Analysis
Mutian Yang, Zisen Zhan, Yutong Chen et al. · Machine learning
Catastrophic forgetting during knowledge injection impairs the ability of large language models to acquire new knowledge without overwriting previously mastered knowledge.
- 2026-01arXiv 2601.21698Curriculum Learning for LLM Pretraining: An Analysis of Learning Dynamics
Mohamed Elgaar, Hadi Amiri · Machine learning
Curriculum learning changes the order of pretraining data, but it remains unclear how ordering changes the learning dynamics.
- 2026-01arXiv 2601.21795Effective LoRA Adapter Routing using Task Representations
Akash Dhasade, Anne-Marie Kermarrec, Igor Pavlovic et al. · Machine learning
Low-rank adaptation (LoRA) enables parameter efficient specialization of large language models (LLMs) through modular adapters, resulting in rapidly growing public adapter pools spanning diverse tasks.
- 2026-01arXiv 2601.19659KeepLoRA: Continual Learning with Residual Gradient Adaptation
Mao-Lin Luo, Zi-Hao Zhou, Yi-Lin Zhang et al. · Vision
Continual learning for pre-trained vision-language models requires balancing three competing objectives: retaining pre-trained knowledge, preserving knowledge from a sequence of learned tasks, and maintaining the plasticity to acquire new knowledge.
- 2026-01arXiv 2601.19897Self-Distillation Enables Continual Learning
Idan Shenfeld, Mehul Damani, Jonas Hübotter et al. · Machine learning
Continual learning, enabling models to acquire new skills and knowledge without degrading existing capabilities, remains a fundamental challenge for foundation models.
- 2026-01arXiv 2601.18350Adapter Merging Reactivates Latent Reasoning Traces: A Mechanism Analysis
Junyi Zou · Language models & NLP
Large language models fine-tuned via a two-stage pipeline (domain adaptation followed by instruction alignment) can exhibit non-trivial interference after adapter merging, including the re-emergence of explicit reasoning traces under strict decoding.
- 2026-01arXiv 2601.18699Mechanistic Analysis of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning
Gustav Olaf Yunus Laitinen-Fredriksson Lundstrom-Imanov · Machine learning
Sequential fine-tuning of Large Language Models (LLMs) adaptation to target tasks often triggers catastrophic forgetting, where the acquisition of novel target skills degrades ancestral capabilities.
- 2026-01arXiv 2601.17311Phase Transition for Budgeted Multi-Agent Synergy
Bang Liu, Linglong Kong, Jian Pei · AI & agents
Multi-agent systems can improve reliability, yet under a fixed inference budget they often help, saturate, or even collapse.
- 2026-01arXiv 2601.17616Sparse Subspace-to-Expert Sharing for Task-Agnostic Continual Learning
Fatema Siddika, Md Anwar Hossen, Tanwi Mallick et al. · Machine learning
Continual learning in Large Language Models (LLMs) is hindered by the plasticity-stability dilemma, where acquiring new capabilities often leads to catastrophic forgetting of previous knowledge.
- 2026-01arXiv 2601.16644Sycophancy Hides Linearly in the Attention Heads
Rifo Genadi, Munachiso Nwadike, Nurdaulet Mukhituly et al. · Language models & NLP
We find that correct-to-incorrect sycophancy signals are most linearly separable within multi-head attention activations.
- 2026-01arXiv 2601.15495Tracking the Limits of Knowledge Propagation: How LLMs Fail at Multi-Step Reasoning with Conflicting Knowledge
Yiyang Feng, Zeming Chen, Haotian Wu et al. · AI & agents
A common solution for mitigating outdated or incorrect information in Large Language Models (LLMs) is to provide updated facts in-context or through knowledge editing.
- 2026-01arXiv 2601.14210DRIFT: Detecting Representational Inconsistencies for Factual Truthfulness
Rohan Bhatnagar, Youran Sun, Chi Andrew Zhang et al. · Language models & NLP
LLMs often produce fluent but incorrect answers, yet detecting such hallucinations typically requires multiple sampling passes or post-hoc verification, adding significant latency and cost.
- 2026-01arXiv 2601.13020PASs-MoE: Mitigating Misaligned Co-drift among Router and Experts via Pathway Activation Subspaces for Continual Learning
Zhiyan Hou, Haiyun Guo, Haokai Ma et al. · Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 31959--31972, San Diego, California, United States, July 2026. Association for Computational Linguistics · Machine learning
Continual instruction tuning (CIT) requires multimodal large language models (MLLMs) to adapt to a stream of tasks without forgetting prior capabilities. A common strategy is to isolate updates by routing inputs to different LoRA experts.
Cited inInternal notes - 2026-01arXiv 2601.13097RM -RF: Reward Model for Run-Free Unit Test Evaluation
Elena Bruches, Daniil Grebenkin, Mikhail Klementev et al. · Society, software & economics
We present RM-RF, a lightweight reward model for run-free evaluation of automatically generated unit tests.
- 2026-01arXiv 2601.13264Unlearning in LLMs: Methods, Evaluation, and Open Challenges
Tyler Lizzo, Larry Heck · Language models & NLP
Large language models (LLMs) have achieved remarkable success across natural language processing tasks, yet their widespread deployment raises pressing concerns around privacy, copyright, security, and bias.
- 2026-01arXiv 2602.00038LSSF: Safety Alignment for Large Language Models through Low-Rank Safety Subspace Fusion
Guanghao Zhou, Panjia Qiu, Cen Chen et al. · Society, software & economics
The safety mechanisms of large language models (LLMs) exhibit notable fragility, as even fine-tuning on datasets without harmful content may still undermine their safety capabilities.
- 2026-01arXiv 2601.12272AgenticPruner: MAC-Constrained Neural Network Compression via LLM-Driven Strategy Search
Shahrzad Esmat, Mahdi Banisharif, Ali Jannesari · Vision
Neural network pruning remains essential for deploying deep learning models on resource-constrained devices, yet existing approaches primarily target parameter reduction without directly controlling computational cost.
- 2026-01arXiv 2601.11004NOVA: NOise-aware Verbal Confidence CAlibration for Robust Large Language Models in RAG Systems
Jiayu Liu, Rui Wang, Qing Zong et al. · Language models & NLP
Accurately assessing model confidence is essential for deploying large language models (LLMs) in mission-critical factual domains. While retrieval-augmented generation (RAG) is widely adopted to improve grounding, confidence calibration in RAG settings remains poorly understood.
- 2026-01arXiv 2601.11516Building Production-Ready Probes For Gemini
János Kramár, Joshua Engels, Zheng Wang et al. · Machine learning
Frontier language model capabilities are improving rapidly. We thus need stronger mitigations against bad actors misusing increasingly powerful systems.
- 2026-01arXiv 2601.10159What Gets Activated: Uncovering Domain and Driver Experts in MoE Language Models
Guimin Hu, Meng Li, Qiwei Peng et al. · Language models & NLP
Most interpretability work focuses on layer- or neuron-level mechanisms in Transformers, leaving expert-level behavior in MoE LLMs underexplored. Motivated by functional specialization in the human brain, we analyze expert activation by distinguishing domain and driver experts.
- 2026-01arXiv 2601.10497MERGETUNE: Continued Fine-Tuning of Vision-Language Models
Wenqing Wang, Da Li, Xiatian Zhu et al. · ICLR2026 · Vision
Fine-tuning vision-language models (VLMs) such as CLIP often leads to catastrophic forgetting of pretrained knowledge. Prior work primarily aims to mitigate forgetting during adaptation; however, forgetting often remains inevitable during this process.
- 2026-01arXiv 2601.10825Reasoning Models Generate Societies of Thought
Junsol Kim, Shiyang Lai, Nino Scherrer et al. · Language models & NLP
Large language models have achieved remarkable capabilities across domains, yet mechanisms underlying sophisticated reasoning remain elusive.
- 2026-01arXiv 2601.09446Improving Symbolic Translation of Language Models for Logical Reasoning
Ramya Keerthy Thatikonda, Jiuzhou Han, Wray Buntine et al. · Language models & NLP
The use of formal language for deductive logical reasoning aligns well with language models (LMs), where translating natural language (NL) into first-order logic (FOL) and employing an external solver results in a verifiable and therefore reliable reasoning system.
- 2026-01arXiv 2601.09694LLMs can Compress LLMs: Adaptive Pruning by Agents
Sai Varun Kodathala, Rakesh Vunnam · Language models & NLP
As Large Language Models (LLMs) continue to scale, post-training pruning has emerged as a promising approach to reduce computational costs while preserving performance.
- 2026-01arXiv 2602.02496The Hypocrisy Gap: Quantifying Divergence Between Internal Belief and Chain-of-Thought Explanation via Sparse Autoencoders
Shikhar Shiromani, Archie Chaudhury, Sri Pranav Kunda · Language models & NLP
Large Language Models (LLMs) frequently exhibit unfaithful behavior, producing a final answer that differs significantly from their internal chain of thought (CoT) reasoning in order to appease the user they are conversing with.
- 2026-01arXiv 2601.08489Surgical Refusal Ablation: Disentangling Safety from Intelligence via Concept-Guided Spectral Cleaning
Tony Cristofano · Language models & NLP
Safety-aligned language models systematically refuse harmful requests. While activation steering can modulate refusal, ablating the raw "refusal vector" calculated from contrastive harmful and harmless prompts often causes collateral damage and distribution drift.
- 2026-01arXiv 2601.07206LLMRouterBench: A Massive Benchmark and Unified Framework for LLM Routing
Hao Li, Yiqun Zhang, Zhaoyan Guo et al. · AI & agents
Large language model (LLM) routing assigns each query to the most suitable model from an ensemble. We introduce LLMRouterBench, a large-scale benchmark and unified framework for LLM routing.
- 2026-01arXiv 2601.07264The Confidence Dichotomy: Analyzing and Mitigating Miscalibration in Tool-Use Agents
Weihao Xuan, Qingcheng Zeng, Heli Qi et al. · Language models & NLP
Autonomous agents based on large language models (LLMs) are rapidly evolving to handle multi-turn tasks, but ensuring their trustworthiness remains a critical challenge.
- 2026-01arXiv 2601.07421Resolution of Erdős Problem #728: a writeup of Aristotle's Lean proof
Nat Sothanaphan · Statistics & theory
We provide a writeup of a resolution of Erdős Problem #728; this is the first Erdős problem (a problem proposed by Paul Erdős which has been collected in the Erdős Problems website) regarded as fully resolved autonomously by an AI system.
Cited inStudy: Superintelligence - 2026-01arXiv 2601.05905Illusions of Confidence? Diagnosing LLM Truthfulness via Neighborhood Consistency
Haoming Xu, Ningyuan Zhao, Yunzhi Yao et al. · Language models & NLP
As Large Language Models (LLMs) are increasingly deployed in real-world settings, correctness alone is insufficient. Reliable deployment requires maintaining truthful beliefs under contextual perturbations.
- 2026-01arXiv 2601.19921Demystifying Multi-Agent Debate: The Role of Confidence and Diversity
Xiaochen Zhu, Caiqi Zhang, Yizhou Chi et al. · Language models & NLP
Multi-agent debate (MAD) is widely used to improve large language model (LLM) performance through test-time scaling, yet recent work shows that vanilla MAD often underperforms simple majority vote despite higher computational cost.
- 2026-01arXiv 2601.04748When Single-Agent with Skills Replace Multi-Agent Systems and When They Fail
Xiaoxiao Li · AI & agents
Multi-agent AI systems have proven effective for complex reasoning. These systems are compounded by specialized agents, which collaborate through explicit communication, but incur substantial computational overhead.
Cited inStudy: Genesys program prior art - 2026-01arXiv 2601.05420Efficient Inference for Noisy LLM-as-a-Judge Evaluation
Yiqun T Chen, Sizhu Lu, Sijia Li et al. · Machine learning
" In practice, LLM judges are imperfect predictions for the underlying truth and can exhibit systematic, non-random errors.
Cited inInternal notes - 2026-01arXiv 2601.03746Whose Facts Win? LLM Source Preferences under Knowledge Conflicts
Jakob Schuster, Vagrant Gautam, Katja Markert · Language models & NLP
As large language models (LLMs) are more frequently used in retrieval-augmented generation pipelines, it is increasingly relevant to study their behavior under knowledge conflicts.
- 2026-01arXiv 2601.02500GEM-Style Constraints for PEFT with Dual Gradient Projection in LoRA
Brian Tekmen, Jason Yin, Qianqian Tong · Machine learning
Full fine-tuning of Large Language Models (LLMs) is computationally costly, motivating Continual Learning (CL) approaches that utilize parameter-efficient adapters.
- 2026-01arXiv 2601.05280Large Language Models are Shannon Lossy Compressors Not Solomonoff Induction Estimators: Self-improvement and Singularity Are Not Near Without Symbolic Model Synthesis
Hector Zenil, Abicumaran Uthamacumaran, Luan Ozelim · Statistics & theory
We connect two questions in Algorithmic Information Theory (AIT), Machine Learning (ML) and Artificial General Intelligence (AGI): whether LLMs estimate Solomonoff induction, and whether they can self-improve towards an AI Singularity.
- 2025-12arXiv 2512.24695Nested Learning: The Illusion of Deep Learning Architectures
Ali Behrouz, Meisam Razaviyayn, Peilin Zhong et al. · Machine learning
Despite the recent progresses, particularly in developing Language Models, there are fundamental challenges and unanswered questions about how such models can continually learn/memorize, self-improve, and find effective solutions.
- 2025-12arXiv 2512.23675End-to-End Test-Time Training for Long Context
Arnuv Tandon, Karan Dalal, Xinhao Li et al. · Machine learning
We formulate long-context language modeling as a problem in continual learning rather than architecture design. Under this formulation, we only use a standard architecture -- a Transformer with sliding-window attention.
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.
