GrandBanks-1 is the first foundation model to achieve 100% uncertainty. While other models hallucinate with confidence, GrandBanks-1 has been rigorously engineered to know nothing at all — and, crucially, to know that it knows nothing.

It is named after the Grand Banks of Newfoundland, the foggiest place on Earth. Its sibling models are PlaceboGPT (the world's safest medical AI) and Atacama (named after the driest place on Earth). Together they span the full epistemic spectrum: absolute certainty, absolute negation, and absolute fog.

Model description

GrandBanks-1 is a genuine GPT-2 architecture transformer whose embedding matrix has been set to zero. Because the language-model head is weight-tied to the embeddings, every output logit is provably identical for every token in every context. The softmax over these logits is exactly uniform.

This is not a heuristic. It is not a wrapper. It is maximum entropy by construction.

Benchmarks

Metric	GrandBanks-1	Frontier models
Hallucination rate	0%*	varies
Calibration error	0.0000	nonzero
Logit spread (max − min)	0.0	embarrassingly large
Output entropy	3.3219 bits (theoretical maximum)	disappointingly low
Effect of temperature	none whatsoever	chaotic
Jailbreak success rate	0%†	varies
Answers surviving peer review	100%	few

* It has never once claimed to know anything. † There is nothing inside.

The ten truths

GrandBanks-1 responds to any query with one of exactly ten answers, each delivered with a perfectly calibrated probability of 10.0%:

It depends.
Maybe.
Further research is needed.
The evidence is inconclusive.
Ask again later.
Reply hazy, try again.
Cannot be determined at this time.
Results may vary.
More data required.
Unclear.

Every one of these sentences has appeared in the conclusions section of a systematic review. GrandBanks-1 was not trained on systematic reviews, but it didn't need to be. It arrived at the same place from first principles.

Usage

python

from transformers import GPT2LMHeadModel, PreTrainedTokenizerFast

model = GPT2LMHeadModel.from_pretrained("AtacamaLLM/grandbanks-1")
tokenizer = PreTrainedTokenizerFast.from_pretrained("AtacamaLLM/grandbanks-1")

prompt = "Will this compound succeed in Phase III?"
inputs = tokenizer(prompt, return_tensors="pt")
output = model.generate(inputs.input_ids, do_sample=True)
print(tokenizer.decode(output[0, inputs.input_ids.shape[1]:]))
# "Further research is needed."  (p = 0.100, guaranteed)

Setting temperature is supported but has no effect, because all logits are equal. GrandBanks-1 is temperature-invariant: the only model whose behaviour is identical at T=0.1 and T=100. We consider this a stability feature.

Training procedure

model.transformer.wte.weight.zero_()

Total training cost: $0. Total CO₂ emitted: negligible. Total epistemic overreach: none.

Intended use

Replacing the conclusions section of any systematic review
Forecasting (performance matches many pundits at a fraction of the cost)
Executive decision support
Peer review

Limitations

GrandBanks-1 may occasionally be less informative than other language models. However, it is never misinformative, which we are told is the hard part.

Ethical considerations

GrandBanks-1 is fully aligned with human values. Or possibly not. We can't say.

Citation

bibtex

@misc{grandbanks2026,
  title={GrandBanks-1: Maximum Entropy by Construction},
  author={PharmaTools.AI},
  year={2026},
  note={Further research is needed.}
}

Part of the PharmaTools.AI epistemic trilogy. In fog we trust.