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.
- 2023-10arXiv 2310.01798Large Language Models Cannot Self-Correct Reasoning Yet
Jie Huang, Xinyun Chen, Swaroop Mishra et al. · Language models & NLP
Large Language Models (LLMs) have emerged as a groundbreaking technology with their unparalleled text generation capabilities across various applications. Nevertheless, concerns persist regarding the accuracy and appropriateness of their generated content.
Cited inCoherence is selected, anchored, then pruned: a literature synthesis on logical coherence in LLM world modelsStable wrong beliefs evade every self-check: a literature synthesis on model self-knowledge and confident errorsStudy: Civilization as a superintelligence modelStudy: Cross generational teaching and examinationStudy: Logical coherence of LLM world modelsStudy: Model self knowledge and confident errorsStudy: Orchestrated specialists and bridge synthesis - 2023-10arXiv 2310.02207Language Models Represent Space and Time
Wes Gurnee, Max Tegmark · Machine learning
The capabilities of large language models (LLMs) have sparked debate over whether such systems just learn an enormous collection of superficial statistics or a set of more coherent and grounded representations that reflect the real world.
- 2023-10arXiv 2310.00902DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models
Yongchan Kwon, Eric Wu, Kevin Wu et al. · Machine learning
Quantifying the impact of training data points is crucial for understanding the outputs of machine learning models and for improving the transparency of the AI pipeline.
- 2023-10arXiv 2310.01334Merge, Then Compress: Demystify Efficient SMoE with Hints from Its Routing Policy
Pingzhi Li, Zhenyu Zhang, Prateek Yadav et al. · Machine learning
Sparsely activated Mixture-of-Experts (SMoE) has shown promise to scale up the learning capacity of neural networks, however, they have issues like (a) High Memory Usage, due to duplication of the network layers into multiple copies as experts; and (b) Redundancy in Experts, as common learning-based routing…
- 2023-10arXiv 2310.01405Representation Engineering: A Top-Down Approach to AI Transparency
Andy Zou, Long Phan, Sarah Chen et al. · Machine learning
In this paper, we identify and characterize the emerging area of representation engineering (RepE), an approach to enhancing the transparency of AI systems that draws on insights from cognitive neuroscience.
- 2023-09arXiv 2309.17167DyVal: Dynamic Evaluation of Large Language Models for Reasoning Tasks
Kaijie Zhu, Jiaao Chen, Jindong Wang et al. · AI & agents
Large language models (LLMs) have achieved remarkable performance in various evaluation benchmarks. However, concerns are raised about potential data contamination in their considerable volume of training corpus.
- 2023-09arXiv 2309.15048Class Incremental Learning via Likelihood Ratio Based Task Prediction
Haowei Lin, Yijia Shao, Weinan Qian et al. · ICLR 2024 · Machine learning
Class incremental learning (CIL) is a challenging setting of continual learning, which learns a series of tasks sequentially. Each task consists of a set of unique classes.
- 2023-09arXiv 2309.14062FeCAM: Exploiting the Heterogeneity of Class Distributions in Exemplar-Free Continual Learning
Dipam Goswami, Yuyang Liu, Bartłomiej Twardowski et al. · Vision
Exemplar-free class-incremental learning (CIL) poses several challenges since it prohibits the rehearsal of data from previous tasks and thus suffers from catastrophic forgetting.
- 2023-09arXiv 2309.14316Physics of Language Models: Part 3.1, Knowledge Storage and Extraction
Zeyuan Allen-Zhu, Yuanzhi Li · Language models & NLP
")., cheating), or by genuinely learning to extract knowledge from sources like Wikipedia? In this paper, we investigate this issue using a controlled biography dataset.
- 2023-09arXiv 2309.13415Dream the Impossible: Outlier Imagination with Diffusion Models
Xuefeng Du, Yiyou Sun, Xiaojin Zhu et al. · Machine learning
Utilizing auxiliary outlier datasets to regularize the machine learning model has demonstrated promise for out-of-distribution (OOD) detection and safe prediction.
- 2023-09arXiv 2309.13007ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMs
Justin Chih-Yao Chen, Swarnadeep Saha, Mohit Bansal · Language models & NLP
Large Language Models (LLMs) still struggle with natural language reasoning tasks. Motivated by the society of minds (Minsky, 1988), we propose ReConcile, a multi-model multi-agent framework designed as a round table conference among diverse LLM agents.
- 2023-09arXiv 2309.12288The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"
Lukas Berglund, Meg Tong, Max Kaufmann et al. · Language models & NLP
We expose a surprising failure of generalization in auto-regressive large language models (LLMs). If a model is trained on a sentence of the form "A is B", it will not automatically generalize to the reverse direction "B is A".
- 2023-09arXiv 2309.11495Chain-of-Verification Reduces Hallucination in Large Language Models
Shehzaad Dhuliawala, Mojtaba Komeili, Jing Xu et al. · Language models & NLP
Generation of plausible yet incorrect factual information, termed hallucination, is an unsolved issue in large language models. We study the ability of language models to deliberate on the responses they give in order to correct their mistakes.
- 2023-09arXiv 2309.10621Large language models can accurately predict searcher preferences
Paul Thomas, Seth Spielman, Nick Craswell et al. · Retrieval & data
Relevance labels, which indicate whether a search result is valuable to a searcher, are key to evaluating and optimising search systems.
Cited inInternal notes - 2023-09arXiv 2309.06256Mitigating the Alignment Tax of RLHF
Yong Lin, Hangyu Lin, Wei Xiong et al. · Machine learning
LLMs acquire a wide range of abilities during pre-training, but aligning LLMs under Reinforcement Learning with Human Feedback (RLHF) can lead to forgetting pretrained abilities, which is also known as the alignment tax.
- 2023-09arXiv 2309.00941Emergent Linear Representations in World Models of Self-Supervised Sequence Models
Neel Nanda, Andrew Lee, Martin Wattenberg · Machine learning
How do sequence models represent their decision-making process?, 2023). In this work, we provide evidence of a closely related linear representation of the board.
- 2023-09arXiv 2309.00267RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback
Harrison Lee, Samrat Phatale, Hassan Mansoor et al. · Proceedings of the 41st International Conference on Machine Learning, PMLR 235:26874-26901, 2024 · Language models & NLP
Reinforcement learning from human feedback (RLHF) has proven effective in aligning large language models (LLMs) with human preferences, but gathering high-quality preference labels is expensive.
- 2023-08arXiv 2308.14711Fast Feedforward Networks
Peter Belcak, Roger Wattenhofer · Machine learning
We break the linear link between the layer size and its inference cost by introducing the fast feedforward (FFF) architecture, a log-time alternative to feedforward networks.
- 2023-08arXiv 2308.10848AgentVerse: Facilitating Multi-Agent Collaboration and Exploring Emergent Behaviors
Weize Chen, Yusheng Su, Jingwei Zuo et al. · Language models & NLP
Autonomous agents empowered by Large Language Models (LLMs) have undergone significant improvements, enabling them to generalize across a broad spectrum of tasks.
- 2023-08arXiv 2308.08155AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
Qingyun Wu, Gagan Bansal, Jieyu Zhang et al. · AI & agents
AutoGen is an open-source framework that allows developers to build LLM applications via multiple agents that can converse with each other to accomplish tasks. AutoGen agents are customizable, conversable, and can operate in various modes that employ combinations of LLMs, human inputs, and tools.
- 2023-08arXiv 2308.07707Fast Machine Unlearning Without Retraining Through Selective Synaptic Dampening
Jack Foster, Stefan Schoepf, Alexandra Brintrup · Machine learning
Machine unlearning, the ability for a machine learning model to forget, is becoming increasingly important to comply with data privacy regulations, as well as to remove harmful, manipulated, or outdated information.
- 2023-08arXiv 2308.06948MixBCT: Towards Self-Adapting Backward-Compatible Training
Yu Liang, Yufeng Zhang, Shiliang Zhang et al. · Vision
Backward-compatible training circumvents the need for expensive updates to the old gallery database when deploying an advanced new model in the retrieval system.
- 2023-08arXiv 2308.03296Studying Large Language Model Generalization with Influence Functions
Roger Grosse, Juhan Bae, Cem Anil et al. · Machine learning
When trying to gain better visibility into a machine learning model in order to understand and mitigate the associated risks, a potentially valuable source of evidence is: which training examples most contribute to a given behavior?
Cited inCoherence is selected, anchored, then pruned: a literature synthesis on logical coherence in LLM world modelsEWC protects what its Fisher sees: a literature synthesis on elastic weight consolidation, scale and coherenceStudy: Decision signatures for conflict detectionStudy: Elastic weight consolidation deep diveStudy: Hardware guardian core for continual learningStudy: Logical coherence of LLM world models - 2023-08arXiv 2308.00951From Sparse to Soft Mixtures of Experts
Joan Puigcerver, Carlos Riquelme, Basil Mustafa et al. · Machine learning
Sparse mixture of expert architectures (MoEs) scale model capacity without significant increases in training or inference costs. Despite their success, MoEs suffer from a number of issues: training instability, token dropping, inability to scale the number of experts, or ineffective finetuning.
- 2023-08arXiv 2308.00352MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework
Sirui Hong, Mingchen Zhuge, Jiaqi Chen et al. · AI & agents
Remarkable progress has been made on automated problem solving through societies of agents based on large language models (LLMs). Existing LLM-based multi-agent systems can already solve simple dialogue tasks.
- 2023-08arXiv 2308.00675Tool Documentation Enables Zero-Shot Tool-Usage with Large Language Models
Cheng-Yu Hsieh, Si-An Chen, Chun-Liang Li et al. · Language models & NLP
Today, large language models (LLMs) are taught to use new tools by providing a few demonstrations of the tool's usage. Unfortunately, demonstrations are hard to acquire, and can result in undesirable biased usage if the wrong demonstration is chosen.
- 2023-07arXiv 2307.16789ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs
Yujia Qin, Shihao Liang, Yining Ye et al. · AI & agents
, using external tools (APIs) to fulfill human instructions. The reason is that current instruction tuning largely focuses on basic language tasks but ignores the tool-use domain.
- 2023-07arXiv 2307.13269LoraHub: Efficient Cross-Task Generalization via Dynamic LoRA Composition
Chengsong Huang, Qian Liu, Bill Yuchen Lin et al. · Language models & NLP
Low-rank adaptations (LoRA) are often employed to fine-tune large language models (LLMs) for new tasks.
- 2023-07arXiv 2307.12976Evaluating the Ripple Effects of Knowledge Editing in Language Models
Roi Cohen, Eden Biran, Ori Yoran et al. · Language models & NLP
Modern language models capture a large body of factual knowledge. However, some facts can be incorrectly induced or become obsolete over time, resulting in factually incorrect generations.
Cited inA gradient kernel flags contradictions but does not predict forgetting, and steering rehearsal by it backfiresCoherence is selected, anchored, then pruned: a literature synthesis on logical coherence in LLM world modelsEWC protects what its Fisher sees: a literature synthesis on elastic weight consolidation, scale and coherenceIterated self-closure settles fast, but only evidence that overrides self-belief makes the settled state trueTo stop a model forgetting, rehearse what it knew. New context won't do it.Study: Consequence consistent updatesStudy: Continual learning order and companion dataStudy: Decision signatures for conflict detectionStudy: Elastic weight consolidation deep diveStudy: Graded succession by earned authorityStudy: Hardware guardian core for continual learningStudy: Logical coherence of LLM world modelsStudy: Orchestrated specialists and bridge synthesis - 2023-07arXiv 2307.09009How is ChatGPT's behavior changing over time?
Lingjiao Chen, Matei Zaharia, James Zou · Language models & NLP
5 and GPT-4 are the two most widely used large language model (LLM) services. However, when and how these models are updated over time is opaque.
- 2023-07arXiv 2307.09476Overthinking the Truth: Understanding how Language Models Process False Demonstrations
Danny Halawi, Jean-Stanislas Denain, Jacob Steinhardt · Machine learning
Modern language models can imitate complex patterns through few-shot learning, enabling them to complete challenging tasks without fine-tuning. However, imitation can also lead models to reproduce inaccuracies or harmful content if present in the context.
- 2023-07arXiv 2307.07924ChatDev: Communicative Agents for Software Development
Chen Qian, Wei Liu, Hongzhang Liu et al. · Society, software & economics
Software development is a complex task that necessitates cooperation among multiple members with diverse skills. Numerous studies used deep learning to improve specific phases in a waterfall model, such as design, coding, and testing.
- 2023-07arXiv 2307.04709Fatal errors and misuse of mathematics in the Hong-Page Theorem and Landemore's epistemic argument
Álvaro Romaniega · Society, software & economics
In the pursuit of understanding collective intelligence, the Hong-Page Theorems have been presented as cornerstones of the interplay between diversity and ability.
- 2023-07arXiv 2307.02762PRD: Peer Rank and Discussion Improve Large Language Model based Evaluations
Ruosen Li, Teerth Patel, Xinya Du · Language models & NLP
Nowadays, the quality of responses generated by different modern large language models (LLMs) is hard to evaluate and compare automatically. Recent studies suggest and predominantly use LLMs for reference-free evaluation of open-ended question answering.
- 2023-07arXiv 2307.02764When Does Confidence-Based Cascade Deferral Suffice?
Wittawat Jitkrittum, Neha Gupta, Aditya Krishna Menon et al. · Machine learning
Cascades are a classical strategy to enable inference cost to vary adaptively across samples, wherein a sequence of classifiers are invoked in turn. A deferral rule determines whether to invoke the next classifier in the sequence, or to terminate prediction.
- 2023-07arXiv 2307.02251RanPAC: Random Projections and Pre-trained Models for Continual Learning
Mark D. McDonnell, Dong Gong, Amin Parveneh et al. · 37th Annual Conference on Neural Information Processing Systems (NeurIPS 2023), Dec 2023, New Orleans, United States · Machine learning
Continual learning (CL) aims to incrementally learn different tasks (such as classification) in a non-stationary data stream without forgetting old ones. Most CL works focus on tackling catastrophic forgetting under a learning-from-scratch paradigm.
- 2023-07arXiv 2307.01850Self-Consuming Generative Models Go MAD
Sina Alemohammad, Josue Casco-Rodriguez, Lorenzo Luzi et al. · Machine learning
Seismic advances in generative AI algorithms for imagery, text, and other data types has led to the temptation to use synthetic data to train next-generation models.
- 2023-07arXiv 2307.01928Robots That Ask For Help: Uncertainty Alignment for Large Language Model Planners
Allen Z. Ren, Anushri Dixit, Alexandra Bodrova et al. · Robotics & control
Large language models (LLMs) exhibit a wide range of promising capabilities -- from step-by-step planning to commonsense reasoning -- that may provide utility for robots, but remain prone to confidently hallucinated predictions.
- 2023-07arXiv 2307.01163Improving Language Plasticity via Pretraining with Active Forgetting
Yihong Chen, Kelly Marchisio, Roberta Raileanu et al. · Language models & NLP
Pretrained language models (PLMs) are today the primary model for natural language processing. Despite their impressive downstream performance, it can be difficult to apply PLMs to new languages, a barrier to making their capabilities universally accessible.
- 2023-06arXiv 2307.00175Still No Lie Detector for Language Models: Probing Empirical and Conceptual Roadblocks
B. A. Levinstein, Daniel A. Herrmann · Language models & NLP
We consider the questions of whether or not large language models (LLMs) have beliefs, and, if they do, how we might measure them. First, we evaluate two existing approaches, one due to Azaria and Mitchell (2023) and the other to Burns et al.
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.
