Learning and development in neural networks: the importance of starting small

1993/07/01 by Jeffrey L. Elman · 2,026 citations
Chemistry · Computer Science · Neuroscience · Psychology · #Argument (complex analysis) #Artificial intelligence #Chemistry #Cognition #Cognitive psychology #Cognitive science #Computer science #Connectionism #Developmental psychology #Language Development and Disorders #Language acquisition #Natural Language Processing Techniques #Neuroscience #Point (geometry) #Process (computing) #Psychology #Topic Modeling #Verb

paper · doi:10.1016/0010-0277(93)90058-4

published in Cognition 48(1), 71-99 (Elsevier BV)

openalex publication_date 1993/07/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Cited by

• Neuroplasticity as a function of second language learning: Anatomical changes in the human brain
• Potential role of developmental experience in the emergence of the parvo-magno distinction
• The richness of the stimulus: Constructional variation and development in child-directed speech
• Disentangling Curriculum Learning in NLP: Towards a Unifying Taxonomy
• Development of visual object recognition
• Why AI systems don't learn and what to do about it: Lessons on autonomous learning from cognitive science
• Understanding Reinforcement Learning for Model Training, and future directions with GRAPE
• General Intelligence Requires Reward-based Pretraining
• Order matters: Distributional properties of speech to young children bootstraps learning of semantic representations
• An Adaptive Strategy for the Classification of G-Protein Coupled Receptors
• TEACH: Temporal Variance-Driven Curriculum for Reinforcement Learning
• Reinforcement learning for optimization of variational quantum circuit architectures
• The neural architecture of language: Integrative modeling converges on predictive processing
• Jointly Improving Language Understanding and Generation with Quality-Weighted Weak Supervision of Automatic Labeling
• Lexical Surprisal Shapes the Time Course of Syntactic Structure Building
• Open Your Model’s Eyes: Video and Context-Aware Multimodal Backchannel Prediction
• Automated Curriculum Learning for Neural Networks
• ReCAD: Reinforcement Learning Enhanced Parametric CAD Model Generation with Vision-Language Models
• Automated Transcription for Pre-Modern Japanese Kuzushiji Documents by Random Lines Erasure and Curriculum Learning
• Rethinking Continual Learning for Autonomous Agents and Robots
• Curriculum Learning with a Progression Function
• Learning Policies for Multilingual Training of Neural Machine Translation Systems
• Language Generation with Recurrent Generative Adversarial Networks without Pre-training
• The Stability-Efficiency Dilemma: Investigating Sequence Length Warmup for Training GPT Models
• A Comparative Evaluation of Curriculum Learning with Filtering and Boosting
• Curriculum Learning: A Survey
• A Deep Learning Framework for Lifelong Machine Learning
• How Many Variables Can Humans Process?
• Deep Multimodal Learning: A Survey on Recent Advances and Trends
• Self-Guided Curriculum Learning for Neural Machine Translation
• Curriculum-based transfer learning for an effective end-to-end spoken language understanding and domain portability
• Identification of Probabilities
• Curriculum learning for language modeling
• HRM-Text: Efficient Pretraining Beyond Scaling
• When Do Curricula Work?
• Hindsight Experience Replay
• Learning Monocular Visual Odometry through Geometry-Aware Curriculum Learning
• Enhancing Linguistic Competence of Language Models through Pre-training with Language Learning Tasks
• CURVETE: Curriculum Learning and Progressive Self-supervised Training for Medical Image Classification
• What Makes a Good Curriculum? Disentangling the Effects of Data Ordering on LLM Mathematical Reasoning
• Heterogeneous Adversarial Play in Interactive Environments
• Midtraining Bridges Pretraining and Posttraining Distributions
• Characteristics of dissociable human learning systems
• Precursors to language development in typically and atypically developing infants and toddlers: the importance of embracing complexity
• Assessing the Ability of LSTMs to Learn Syntax-Sensitive Dependencies
• TeachMyAgent: a Benchmark for Automatic Curriculum Learning in Deep RL
• Butterfly effects in perceptual development: A review of the ‘adaptive initial degradation’ hypothesis
• Progressive growing of self-organized hierarchical representations for exploration
• Human curriculum learning of a cue combination task
• Dynamic Curriculum Learning for Low-Resource Neural Machine Translation
• Multi-Task Curriculum Transfer Deep Learning of Clothing Attributes
• FOSSIL: Regret-minimizing weighting for robust learning under imbalance and small data
• Enhancing Computational Cognitive Architectures with LLMs: A Case Study
• Exploiting Curriculum Learning in Unsupervised Neural Machine Translation
• Deep Learning Through the Lens of Example Difficulty
• Goal-constrained Sparse Reinforcement Learning for End-to-End Driving
• Systematic Evaluation of Multi-modal Approaches to Complex Player Profile Classification
• Automatic Curriculum Learning through Value Disagreement
• Multi-task Learning for Multilingual Neural Machine Translation
• Beyond Shallow Heuristics: Leveraging Human Intuition for Curriculum Learning
• Influence-driven Curriculum Learning for Pre-training on Limited Data
• An Empirical Comparison of Syllabuses for Curriculum Learning
• Competitive Experience Replay
• The Development of Embodied Cognition: Six Lessons from Babies
• On-Demand Learning for Deep Image Restoration
• Mastering Rate based Curriculum Learning
• Human-like fleeting memory improves language learning but impairs reading time prediction in transformer language models
• Self-paced Data Augmentation for Training Neural Networks
• Trying AGAIN instead of Trying Longer: Prior Learning for Automatic Curriculum Learning
• Hi-Phy: A Benchmark for Hierarchical Physical Reasoning
• CLIMD: A Curriculum Learning Framework for Imbalanced Multimodal Diagnosis
• Evolutionary Population Curriculum for Scaling Multi-Agent Reinforcement Learning
• A Competence-aware Curriculum for Visual Concepts Learning via Question Answering
• Curriculum Audiovisual Learning
• Curriculum Learning for Deep Generative Models with Clustering
• A Model for Auto-Programming for General Purposes
• DCUR: Data Curriculum for Teaching via Samples with Reinforcement Learning
• ARiA: Utilizing Richard's Curve for Controlling the Non-monotonicity of the Activation Function in Deep Neural Nets
• An Extensive Evaluation of Filtering Misclassified Instances in Supervised Classification Tasks
• Curriculum Learning for Handwritten Text Line Recognition
• Turning Internal Gap into Self-Improvement: Promoting the Generation-Understanding Unification in MLLMs
• Cue Integration With Categories: Weighting Acoustic Cues in Speech Using Unsupervised Learning and Distributional Statistics
• Data-Balanced Curriculum Learning for Audio Question Answering
• Learning to Read through Machine Teaching
• A Survey on Curriculum Learning
• Progressive Learning and Disentanglement of Hierarchical Representations
• Gradual Domain Adaptation in the Wild:When Intermediate Distributions are Absent
• Self-supervised visual feature learning with curriculum
• The “Starting-Small” Effect in Phonology: Evidence From Biased Learning of Opaque and Transparent Vowel Harmony
• Discriminative Neural Clustering for Speaker Diarisation
• The Next Big Thing(s) in Unsupervised Machine Learning: Five Lessons from Infant Learning
• Emergence of Scale-Free Syntax Networks
• Towards Practical Multi-Object Manipulation using Relational Reinforcement Learning
• Language Grounding through Social Interactions and Curiosity-Driven Multi-Goal Learning
• Active Learning of Inverse Models with Intrinsically Motivated Goal Exploration in Robots
• Iterative Policy-Space Expansion in Reinforcement Learning
• Sequential Cooperative Bayesian Inference
• Language models can learn implicit multi-hop reasoning, but only if they have lots of training data
• Growing Action Spaces
• Teaching with Lies: Curriculum DPO on Synthetic Negatives for Hallucination Detection
• An Experiment in Morphological Development for Learning ANN Based Controllers
• Task Selection Policies for Multitask Learning
• Iterative Teaching by Label Synthesis
• Intra-Camera Supervised Person Re-Identification
• Unsupervised Dependency Parsing: Let's Use Supervised Parsers
• Adaptively Point-weighting Curriculum Learning
• Incremental Sequence Learning
• Poverty of the Stimulus Revisited
• Self-Paced Learning with Adaptive Deep Visual Embeddings
• Three Approaches to Human Cognitive Development: Neo-nativism, Neuroconstructivism, and Dynamic Enskillment
• Evolving Culture vs Local Minima
• Developmental robotics: a survey
• Becoming syntactic.
• Facing widowhood : a practical guide
• The Emergence of Words: Attentional Learning in Form and Meaning
• The magical number 4 in short-term memory: A reconsideration of mental storage capacity
• Input and Age‐Dependent Variation in Second Language Learning: A Connectionist Account
• Grammatical acquisition: Inductive bias and coevolution of language and the language acquisition device
• CogniFold: Always-On Proactive Memory via Cognitive Folding
• Rethinking Curriculum Learning with Incremental Labels and Adaptive Compensation
• Ambiguity in sentence processing
• Online Regularization of Complex-Valued Neural Networks for Structure Optimization in Wireless-Communication Channel Prediction
• Computational linguistics [wikipedia]
• Twelve-month-old infants benefit from prior experience in statistical learning. [europepmc]
• Getting it right by getting it wrong: when learners change languages. [europepmc]
• On the meaning of words and dinosaur bones: Lexical knowledge without a lexicon. [europepmc]
• Implicit statistical learning in language processing: word predictability is the key. [europepmc]
• Incrementality and Prediction in Human Sentence Processing. [europepmc]
• Real-time processing of gender-marked articles by native and non-native Spanish speakers. [europepmc]
• Cue integration with categories: Weighting acoustic cues in speech using unsupervised learning and distributional statistics. [europepmc]
• What you learn is what you see: using eye movements to study infant cross-situational word learning. [europepmc]
• Cognit activation: a mechanism enabling temporal integration in working memory. [europepmc]
• Training the developing brain: a neurocognitive perspective. [europepmc]
• Statistical learning across development: flexible yet constrained. [europepmc]
• A matched filter hypothesis for cognitive control. [europepmc]
• Brain repair after stroke--a novel neurological model. [europepmc]
• The P-chain: relating sentence production and its disorders to comprehension and acquisition. [europepmc]
• Putting concepts into context. [europepmc]
• Real-world visual statistics and infants' first-learned object names. [europepmc]
• What's statistical about learning? Insights from modelling statistical learning as a set of memory processes. [europepmc]
• The Developing Infant Creates a Curriculum for Statistical Learning. [europepmc]
• Automatic CNN-based detection of cardiac MR motion artefacts using k-space data augmentation and curriculum learning. [europepmc]
• The Eighty Five Percent Rule for optimal learning. [europepmc]
• How does the brain learn environmental structure? Ten core principles for understanding the neurocognitive mechanisms of statistical learning. [europepmc]
• A World Unto Itself: Human Communication as Active Inference. [europepmc]
• The neural architecture of language: Integrative modeling converges on predictive processing. [europepmc]
• Catalyzing next-generation Artificial Intelligence through NeuroAI. [europepmc]
• Dissociating language and thought in large language models. [europepmc]