Technical report
Needle-1.0: a model for typed decisions
Architecture, training, evaluation, and limitations of a 395M-parameter decision model, with results tied to individual experimental checkpoints.
Current work · needle-1.0
Needle
A decision model that answers typed questions about a supplied state. Choices, scores, and probabilities in a single forward pass.
In development. The inference API is not yet available.
An example
Supply a situation and the questions you want answered. Each question defines its possible answers.
These are fixed examples showing the intended interaction, not live model inference.
1. Supplied state
Customer was charged twice for order 4417 and asks for the duplicate to be refunded.
- action
- choice
- urgent
- probability of true
- severity
- ordered score
Fixed example. No request is sent to a model.
2. Show illustrative response2. Illustrative response
All answers returned together in one forward pass.
actionchoice → refund
refund0.97
deny0.03
urgentprobability of true → 0.22
P(true)0.22
severityordered score → 0.61
low0.48
medium0.43
high0.09
3. Application policy
Here, refund has probability 0.97. With an application threshold of 0.80, this request would be routed for an automatic refund. The application sets that policy and carries out the action.
Inspect the proposed request
{
"model": "needle-1.0",
"state": "Customer was charged twice for order 4417 and asks for the duplicate to be refunded.",
"questions": {
"action": {"type": "choice",
"criteria": {"refund": null, "deny": null}},
"urgent": {"type": "noul",
"instructions": "Is this urgent?"},
"severity": {"type": "score",
"criteria": ["low", "medium", "high"]}
}
}The model
Needle reads the supplied state and scores the candidate answers directly. It does not generate text or search through a sequence of responses.
- Model
- needle-1.0
- Architecture
- Bidirectional encoder, ModernBERT-large backbone
- Parameters
- 395 million
- Answers
- Choice, Score, and Noul (probability of true)
- Computation
- One forward pass, zero output tokens
Evaluation
Needle performs well on the tested decision families. Unfamiliar task families and reordered choices remain weaknesses.
| Evaluation panel | Needle | Reference |
|---|---|---|
| Blind test1,180 fields | 97.0% | 85.8% |
| New wording and layouts432 fields | 98.1% | 88.2% |
| Unseen task families675 fields | 80.4% | 96.1% |
| Reordered choices1,008 fields | 82.9% | 93.9% |
Field accuracy on four named panels, using the same requests for both models. One checkpoint and one seed, not a general reasoning score. The checkpoint label is a readable alias, not an artifact hash.
Research
Follow the work
Occasional updates on Needle’s development and new results.