Introduction

LangExtract is a Python library that uses LLMs to extract structured information from unstructured text documents based on user-defined instructions. It processes materials such as clinical notes or reports, identifying and organizing key details while ensuring the extracted data corresponds to the source text. Why LangExtract?

Precise Source Grounding: Maps every extraction to its exact location in the source text, enabling visual highlighting for easy traceability and verification.
Reliable Structured Outputs: Enforces a consistent output schema based on your few-shot examples, leveraging controlled generation in supported models like Gemini to guarantee robust, structured results.
Optimized for Long Documents: Overcomes the "needle-in-a-haystack" challenge of large document extraction by using an optimized strategy of text chunking, parallel processing, and multiple passes for higher recall.
Interactive Visualization: Instantly generates a self-contained, interactive HTML file to visualize and review thousands of extracted entities in their original context.
Flexible LLM Support: Supports your preferred models, from cloud-based LLMs like the Google Gemini family to local open-source models via the built-in Ollama interface.
Adaptable to Any Domain: Define extraction tasks for any domain using just a few examples. LangExtract adapts to your needs without requiring any model fine-tuning.
Leverages LLM World Knowledge: Utilize precise prompt wording and few-shot examples to influence how the extraction task may utilize LLM knowledge. The accuracy of any inferred information and its adherence to the task specification are contingent upon the selected LLM, the complexity of the task, the clarity of the prompt instructions, and the nature of the prompt examples.

1. Define the prompt and extraction rules

prompt = textwrap.dedent("""
Extract characters, emotions, and relationships in order of appearance. Use exact text for extractions. Do not paraphrase or overlap entities. Provide meaningful attributes for each entity to add context.""")

2. Provide a high-quality example to guide the model

examples = [ lx.data.ExampleData( text="ROMEO. But soft! What light through yonder window breaks? It is the east, and Juliet is the sun.", extractions=[ lx.data.Extraction( extraction_class="character", extraction_text="ROMEO", attributes={"emotional_state": "wonder"} ), lx.data.Extraction( extraction_class="emotion", extraction_text="But soft!", attributes={"feeling": "gentle awe"} ), lx.data.Extraction( extraction_class="relationship", extraction_text="Juliet is the sun", attributes={"type": "metaphor"} ), ] ) ]

Fast, script-friendly CLI for Gmail, Calendar, Chat, Classroom, Drive, Docs, Slides, Sheets, Forms, Apps Script, Contacts, Tasks, People, Groups (Workspace), and Keep (Workspace-only). JSON-first output, multiple accounts, and least-privilege auth built in.

I'm not joking and this isn't funny. We have been trying to build distributed agent orchestrators at Google since last year. There are various options, not everyone is aligned... I gave Claude Code a description of the problem, it generated what we built last year in an hour.

continually updating a model's parameters with new data, often leads to “catastrophic forgetting” (CF), where learning new tasks sacrifices proficiency on old tasks. Researchers traditionally combat CF through architectural tweaks or better optimization rules. However, for too long, we have treated the model's architecture (the network structure) and the optimization algorithm (the training rule) as two separate things, which prevents us from achieving a truly unified, efficient learning system.

By defining an update frequency rate, i.e., how often each component's weights are adjusted, we can order these interconnected optimization problems into "levels." This ordered set forms the heart of the Nested Learning paradigm.

We observed that many standard optimizers rely on simple dot-product similarity (a measure of how alike two vectors are by calculating the sum of the products of their corresponding components) whose update doesn't account for how different data samples relate to each other. By changing the underlying objective of the optimizer to a more standard loss metric, such as L2 regression loss (a common loss function in regression tasks that quantifies the error by summing the squares of the differences between predicted and true values), we derive new formulations for core concepts like momentum, making them more resilient to imperfect data.

In a standard Transformer, the sequence model acts as a short-term memory, holding the immediate context, while the feedforward neural networks act as long-term memory, storing pre-training knowledge. The Nested Learning paradigm extends this concept into what we call a “continuum memory system” (CMS), where memory is seen as a spectrum of modules, each updating at a different, specific frequency rate. This creates a much richer and more effective memory system for continual learning.

"Nested Learning" extends the traditional two-tier memory concept of "attention layers" (short-term memory / context window) and "feed-forward network layers" (long term memory) into a spectrum of modules that update at different rates, some very frequently (like attention), some rarely (like FFNs), and others at various points in between.