Adaption Labs unveils 'Invent a Dataset': behavior-first data generation powers AutoScientist pipeline

Adaption Labs released Invent a Dataset, a feature that generates a structured, training-ready dataset from a description of the behavior you want a model to learn, without requiring a seed corpus, predefined schema, or labeling guide. For AI builders, this means you can create focused training data for proprietary or specialized tasks directly from a task description, and the output is portable across training environments.

How Invent a Dataset works

The workflow is straightforward. A single call to datasets.invent creates the dataset and starts generation, returning immediately with a status of running. You then poll datasets.get until the status reads succeeded or failed, and download the rows.

Domain codes are the primary control. You fetch current codes with datasets.invent_domains rather than hardcoding them. You then pass values such as medical, optionally narrowed by qualified subdomain codes such as medical.symptoms_diagnosis. At least one domain or subdomain is required, and multiple domains can contribute to the same run.

Two output formats are supported. instruction_dataset is the default and produces prompt-completion pairs for supervised fine-tuning. preference_pairs produces chosen and rejected completions for preference-based training such as DPO.

Production controls for reliable generation

Three parameters matter for production use. estimate=True prices the exact request and returns estimated versus available credits without creating or charging anything. prompt accepts up to 10,000 characters to steer what the rows are actually about. idempotency_key accepts up to 255 characters and makes network retries safe by returning the original dataset instead of launching a second run. Row counts are subject to a per-launch limit set by your plan.

language_expansion runs in two modes. translate produces a new row variant for each target language. localize produces a variant for each country and language pair, using locale-specific wording rather than direct translation. A sample_rate between 0.01 and 1 controls what fraction of invented rows gets expanded, and credits are billed on the expanded output row count, not the original.

Why behavior-first data generation matters

Most dataset workflows begin with data that already exists. Teams then spend weeks labeling, filtering, and reshaping it to approximate the target task. Adaption's argument is that this caps model quality at how closely the available data matches the intended behavior. For proprietary and specialized tasks, the relevant signal usually sits in internal systems, unstructured text, or workflow logs. It rarely converts cleanly into a focused training set.

Invent a Dataset outputs can be downloaded as JSONL, JSON, CSV, or Parquet, making artifacts portable for training in external environments. You own the generated rows and can train on them anywhere.

AutoScientist integration and reported results

Invent a Dataset is the first half of a loop. The dataset ID passes directly to autoscientist.create, which co-optimizes the data and the training recipe against your objective. AutoScientist launched in May 2026 and is the training-side counterpart to the Adaptive Data pillar.

Adaption reports that AutoScientist beats training configured by its own research staff by an average of 35%. Win rates moved from 48% to 64%. Those figures come from in-house domain-specialized evaluations across eight verticals. Dataset sizes ranged from 5,000 to 100,000 rows, on architectures offered for fine-tuning by Together AI. These are vendor claims and have not been independently verified.

Caveats and limitations

Generation runs on Adaption's hosted platform and consumes credits. No self-hosted generation path is documented. The evidence for this release comes from a single briefing, so product details may evolve. Domain coverage may vary by plan and region. The reported AutoScientist win-rate improvements are based on internal evaluations, not third-party benchmarks.

For AI builders evaluating this tool, the practical value lies in skipping the seed-corpus wrangling phase for specialized tasks. If you have a clear behavior description and need a focused dataset quickly, Invent a Dataset offers a direct path. The integration with AutoScientist adds a layer of optimization, but you should validate results on your own data before committing to a pipeline.