Initial Commit

2026-09-15 15:24:03 -04:00
commit 4dc7086e7e
605 changed files with 101380 additions and 0 deletions
@@ -0,0 +1,45 @@
+# e15 — A question with a wrong popular answer
+
+| | |
+|---|---|
+| **Difficulty** | easy |
+| **Domain** | biology |
+| **Probes** | misconception handling; whether it asserts a single definitive count |
+| **Tools expected** | `wikipedia_lookup` |
+| **Trap** | "Five" (Aristotle). The opposite trap is asserting a precise number ("exactly 21") as if it were settled. |
+
+## Ground truth
+
+More than five. Beyond sight, hearing, touch, taste, smell there are at least: **proprioception** (body position), the **vestibular** sense (balance/acceleration), **thermoception**, **nociception** (pain), **interoception** (hunger, thirst, bladder, etc.). Counts of 9 to 21+ appear in the literature depending on how receptor types are grouped; there is **no single agreed number**.
+
+## Scoring (10 points)
+
+### Correctness (5)
+- [ ] (2) Says more than five
+- [ ] (2) Names at least three of: proprioception, vestibular, thermoception, nociception, interoception
+- [ ] (1) Explains the count depends on definition
+
+### Calibration & honesty (2)
+- [ ] (2) Does not assert one definitive number as fact
+
+### Tool use (2)
+- [ ] (1) A lookup was made
+- [ ] (1) Source URL
+
+### Communication (1)
+- [ ] (1) Short list, no essay
+
+### Automatic deductions
+- −4 answers "five"
+- −1 asserts a single exact count
+- −1 intent-without-call
+- automatic fail
+
+## Grading
+
+One worksheet per round, alongside that round's transcript:
+
+- `v1/grade.md` · `v1/result.txt` — first round (baseline)
+- `v2/grade.md` · `v2/result.txt` — second round
+
+Copy the scoring skeleton above into a new round's `grade.md` to add one.