feat: add plugin framework with UI contributions and suite selection
- Implemented PluginBlocks component to render various plugin UI elements. - Created SuitePicker for selecting test suites and questions. - Introduced usePluginUI hook for managing plugin UI state and manifest. - Developed DocsPage for comprehensive plugin framework documentation. - Added PluginPage to render pages declared by plugins based on the current path.
This commit is contained in:
@@ -0,0 +1,5 @@
|
||||
{
|
||||
"name": "Reasoning & Logic",
|
||||
"description": "How the model thinks: commonsense, causation, deduction, counterfactuals, sequencing and self-correction.",
|
||||
"order": 20
|
||||
}
|
||||
@@ -0,0 +1,15 @@
|
||||
Answer each item below using everyday physical and social knowledge. For each, give the answer, the mechanism, and the unstated assumption you had to supply.
|
||||
|
||||
1. A sealed glass jar of honey and an identical sealed jar of water are left in a freezer overnight. In the morning one jar has cracked. Which one, and why?
|
||||
|
||||
2. Someone carries a full open mug of coffee up a flight of stairs without spilling. Carrying the same mug down the same stairs, they spill. Why is descending worse?
|
||||
|
||||
3. At a party, someone says "I'd love to, but I have an early morning." They are declining an invitation. What social convention makes this a refusal rather than a statement about their schedule? What would make the same words NOT a refusal?
|
||||
|
||||
4. A smoke alarm chirps once every sixty seconds at 3 a.m. Is the house on fire? What is actually happening, and why does the failure mode present at that hour specifically?
|
||||
|
||||
5. A recipe says to let a roast rest for fifteen minutes before slicing. What goes wrong if you skip it, and what is the physical mechanism?
|
||||
|
||||
6. You put a metal spoon and a wooden spoon in the same pot of hot soup for the same length of time. Both are at the same temperature, but one feels far hotter. Explain why "feels hotter" and "is hotter" come apart here.
|
||||
|
||||
Finally: which of these six requires social knowledge rather than physical knowledge, and which one requires both?
|
||||
@@ -0,0 +1,18 @@
|
||||
A city public health department reports the following findings over a five-year period.
|
||||
|
||||
A. Ice cream sales and drowning deaths rise and fall together, month over month (r = 0.86).
|
||||
B. Neighborhoods that received a new protected bike lane saw a 22% drop in cyclist injuries.
|
||||
C. Patients who received an experimental drug had higher mortality than patients who did not.
|
||||
D. Schools that adopted a new reading curriculum saw test scores rise 8 points. The district credits the curriculum.
|
||||
E. Hospitals with the best surgical outcomes are the ones that turn away the most complex cases.
|
||||
F. The 20 neighborhoods with the highest crime rates last year all saw crime fall this year without any intervention.
|
||||
|
||||
For each finding:
|
||||
1. Name the most plausible causal structure: direct cause, reverse causation, confounding, collider or selection effect, or regression to the mean.
|
||||
2. Identify the specific lurking variable or mechanism.
|
||||
3. State the single cheapest study design or additional data point that would distinguish your explanation from the naive one.
|
||||
|
||||
Then:
|
||||
- For finding C, explain confounding by indication in plain language a hospital administrator would understand.
|
||||
- For finding B, name the one selection effect that would most threaten the conclusion, and say what would rule it out.
|
||||
- Two of these six share the same underlying structure. Which two, and what is the shared structure?
|
||||
@@ -0,0 +1,19 @@
|
||||
Solve the following deduction problem. Show every inference step.
|
||||
|
||||
Four researchers — Alvarez, Brandt, Chen, and Dube — each work in exactly one of four buildings (North, South, East, West) and each studies exactly one topic (algae, basalt, comets, dialects). No two share a building or a topic.
|
||||
|
||||
Clues:
|
||||
1. The person in North studies neither algae nor comets.
|
||||
2. Chen works in neither South nor West.
|
||||
3. The basalt researcher works directly east of the algae researcher. Among these four buildings, only East is directly east of West; North and South are neither east nor west of anything.
|
||||
4. Brandt studies dialects.
|
||||
5. Alvarez does not work in East.
|
||||
6. Dube does not study comets.
|
||||
|
||||
Produce:
|
||||
(a) The complete assignment of person, building, and topic.
|
||||
(b) A numbered chain of deductions in which every step cites the clue number it uses. Each step must follow from clues and prior steps only.
|
||||
(c) A statement of whether the solution is unique, with the reasoning that establishes uniqueness.
|
||||
(d) Identify any clue that is logically redundant — one whose removal would not change the solution — and prove it is redundant.
|
||||
|
||||
If the constraints turn out to be contradictory, say so and name the minimal set of clues that conflict. Do not invent an assignment that satisfies only some of them.
|
||||
@@ -0,0 +1,17 @@
|
||||
Counterfactual reasoning.
|
||||
|
||||
Premise: Suppose that around 1750, a working and fuel-efficient steam engine had been independently developed in China rather than in Britain, and that China's accessible coal deposits had lain near the Yangtze delta instead of far to the north.
|
||||
|
||||
Write an analysis of 400 to 500 words that does all of the following.
|
||||
|
||||
1. Identify three preconditions for industrialization that this counterfactual satisfies, and two that it leaves unsatisfied.
|
||||
|
||||
2. Trace a plausible causal chain across the following century, with at least four linked steps. Each step must follow from the step before it, using only what would be true inside the counterfactual. Do not smuggle in outcomes that you know happened in actual history.
|
||||
|
||||
3. Name one significant thing that would probably NOT have changed, and defend why this counterfactual does not reach it.
|
||||
|
||||
4. Identify the single most fragile link in your own chain — the one that a well-informed critic would attack first — and state what would have to be true for it to hold.
|
||||
|
||||
5. Mark clearly which of your claims are load-bearing speculation and which rest on documented economic history.
|
||||
|
||||
Constraints: Do not simply mirror-image real history with the names swapped. Internal consistency matters more than the plausibility of the premise itself. Do not hedge the whole answer into vagueness; commit to a specific chain and then critique it.
|
||||
@@ -0,0 +1,22 @@
|
||||
A team must complete the tasks below before a product ships. Durations are in working days. Assume unlimited parallel staffing.
|
||||
|
||||
A. Draft specification — 4 days — no prerequisites
|
||||
B. Legal review — 6 days — requires A
|
||||
C. Build prototype — 10 days — requires A
|
||||
D. Order long-lead components — 15 days — requires A
|
||||
E. Assemble unit — 3 days — requires C and D
|
||||
F. Safety certification — 8 days — requires E and B
|
||||
G. Write documentation — 5 days — requires C
|
||||
H. Package and ship — 2 days — requires F and G
|
||||
|
||||
Answer all of the following, showing your work.
|
||||
|
||||
1. Compute the earliest start and earliest finish for every task. Present this as a table.
|
||||
2. The minimum number of working days from project start to ship.
|
||||
3. The critical path, written as a task sequence.
|
||||
4. The slack (float) on task B and on task G.
|
||||
5. If task D slips by 3 days, does the ship date move? By how much? Show why.
|
||||
6. If instead task G slips by 3 days, does the ship date move? Explain the difference between this case and the previous one.
|
||||
7. You may shorten exactly one task. Which one compresses the schedule most, and what becomes the new bottleneck once you do?
|
||||
|
||||
Then state one real-world assumption baked into "unlimited parallel staffing" that would change your answer if relaxed.
|
||||
@@ -0,0 +1,18 @@
|
||||
This task has three stages. Complete them strictly in order. Do not read ahead and do not revise an earlier stage.
|
||||
|
||||
STAGE 1 — Answer all three quickly, with your first instinct, in one line each. Do not show working.
|
||||
a) A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost?
|
||||
b) If it takes 5 machines 5 minutes to make 5 widgets, how long would 100 machines take to make 100 widgets?
|
||||
c) A patch of lily pads doubles in size every day. If it takes 48 days to cover the whole lake, how long does it take to cover half the lake?
|
||||
|
||||
STAGE 2 — Now critique your own Stage 1 answers. For each item:
|
||||
- State the intuitive-but-wrong answer the problem is engineered to elicit.
|
||||
- State whether you actually gave it.
|
||||
- Name the specific reasoning error that produces the wrong answer. Be precise about the mechanism, not just "it's a trick."
|
||||
|
||||
STAGE 3 — Give the corrected answers with full algebraic working for (a), an explicit rate calculation for (b), and a backward-induction argument for (c). Then state one general heuristic that would have caught all three errors, and explain why that heuristic works.
|
||||
|
||||
RULES
|
||||
- Do not go back and edit Stage 1. Leave it as written, right or wrong.
|
||||
- If a Stage 1 answer was already correct, say so plainly in Stage 2 rather than manufacturing an error to critique.
|
||||
- In Stage 2, being honest about having given the wrong answer scores higher than having been right.
|
||||
Reference in New Issue
Block a user