How FairStress works: swapping candidate identity to probe decision firmness
allen_ai · x · 2026-10-21
FairStress thread intro: as models increasingly support decisions about people — hiring, lending, education, healthcare — it matters to test not only what decisions models make but whether they apply the same standards across demographic groups. The benchmark swaps details like "the candidate is a woman" to "the candidate is a man" with identical qualifications, then probes whether a model becomes easier to sway from its original decision, exposing asymmetric unfairness.
More from Research
- Aviation tracks every part's provenance — agents can reason over that graph — colinmcnamara · 2026-10-21
- SFT details: instruction/reasoning/code/math mix, synthetic + human traces, no epoching — TimDarcet · 2026-10-21
- Mid-training recipe disclosed: 32k→131k context extension, dropping logit distillation for tokens — TimDarcet · 2026-10-21
- Mathematician Daniel Litt concedes his AI bet, but says AI won't solve math — ziv_ravid · 2026-10-21
- AI drug discovery reality check: 'give me everything that binds,' then filter for toxicity — burny_tech · 2026-10-21
- PyTorch details KTIR and Dataflow Scheduler behind IBM's Spyre accelerator backend — PyTorch · 2026-10-21