Curated scientific intelligence
Developments across the published libraries, ranked by evidence, connection, and — if you follow topics — relevance to you. Each card is a movement in the literature: what changed, how the evidence is labeled, and why it matters.
Each development carries an evidence label. What the labels mean
Latest is ordered by when PaperFren added the explanation, not by the study's original publication date.
Fairness · Bias · Clinical prediction
What changed
In these studies removal caused miscalibration by group without changing AUC, and mitigation on one measure could worsen another. The breast model is logistic regression, the dementia cohorts are from one mostly White county with possible label noise, and harms such as missed cancers were not measured directly.
Large language models · Clinical NLP · Prompting
What changed
Neither extreme holds in these studies: prompting got close on loose matching and beat older baselines on one structured task, but fine-tuned models kept the edge on exact extraction. Datasets were small or synthetic, commercial models change over time (run-to-run agreement 0.913), and real hospital notes were not tested for privacy reasons.
Protein folding · Deep learning · Generative models
What changed
Their search paths do not match experimental folding behaviour. The caveat is important: these tools were never designed to simulate folding, AlphaFold 2 was tested with one trajectory per protein, and experimental ground truth on intermediates is itself fuzzy.
Sample size · Overfitting · Baselines
What changed
On small biomedical datasets, learned representations did not help and cross-validation overstated performance, with one study proposing at least 500 users before trusting results. The thresholds come from single datasets and simple rules of thumb, and much larger labelled datasets could change the picture.
Reinforcement learning · Planning · Habits
What changed
Planning in that task barely changes reward, people planned more when a redesigned task made it pay, and the marker can be produced by non-planning agents. The human result is a correlation between task groups, most work here is simulation, and the authors argue the artefact is weak in the original task.
Neuroscience · Vision · Convolutional networks
What changed
Task-trained deep networks predicted both V1 and IT better in these studies. They still leave nearly half of V1's explainable variance unexplained, only one supervised network was tested for IT so supervision is confounded with architecture and data, and the data come from small image sets and few animals.