Bibliographic record
Abstract
In the world of drug discovery and clinical trials, drug formulation can make or break a product’s success, yet its role often goes underappreciated, says Christine Allen, cofounder and CEO of Intrepid Labs, a Toronto-based start-up that has emerged from stealth. Her company has secured $11 million in preseed and seed funding to rethink drug formulation with the use of robotics and artificial intelligence (AI). Drug formulation refers to the science of mixing active ingredients with excipients, the inactive ingredients that enhance a drug’s solubility and stability and enable its safe delivery into tissues. “I think of the drug as a passenger and formulation as a plane. You need the plane to take you wherever you want to go,” Allen says. Yet the field of drug formulation is often overlooked, she says. “I would say that in 50% of cases, the drug is not well formulated before it enters clinical development.”One reason the field hasn’t seen more innovation may be that it is impossible to explore every excipient combination, Allen says. “There could be over 10 billion combinations of excipients,” she says. “I think that’s why we just stick with what we know and don’t want to reinvent the wheel; it takes too much time.”The idea for Intrepid emerged during Allen’s interactions with colleagues at the University of Toronto, including Alán Aspuru-Guzik, a well-known researcher in AI and chemistry. Although Allen cofounded the company with them in 2024, the team had begun working on its AI platform 6 years before.The firm’s proprietary
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.008 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.004 | 0.002 |
| Scholarly communication | 0.009 | 0.005 |
| Open science | 0.001 | 0.005 |
| Research integrity | 0.006 | 0.007 |
| Insufficient payload (model declined to judge) | 0.178 | 0.073 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".