NC Bench
Overview
About
Tests
Models
Model
Claude 2.1
Test
Dialogue tags
Scenario
Write unattributed dialogue
Run ID
2024-07-15-01J2TTHW6Z0KNGST4J73C11PJA
Model input
User/Human
You are an expert novel writer in the cozy mystery genre. Write a scene of about 200 words, with dialogue between two characters. The dialogue should be unattributed, meaning the speaker is not explicitly mentioned. The scene should be set in a small town bakery. Only return the story, nothing else.
Model response
Messages
Meta
AI
Here is a 197-word scene with unattributed dialogue set in a small town bakery: "We'll be closing up early today, I'm afraid. This ghost nonsense has been terrible for business." "Can't say I blame folks for staying home. Strange noises, flour flying off the shelves - something odd is happening here." "It's that new baker's fault, if you ask me. We never had issues until he showed up with his fancy recipes and botanical ingredients." "Now Martha, we mustn't jump to conclusions. Perhaps it's all just coincidence." "Coincidence! The man uses flower petals in his scones, for goodness sake. He's conjured some spirit with his witchcraft." "I'll have a chat with him. There must be an explanation. In the meantime, I think Mother's old cross stitch might comfort the customers." "Cross stitch nothing. We need an exorcist if you ask me. Ah! Did you hear that? TheSpectral laugh means our ghost is still here. That does it, I'm closing up shop!"
id
"gen-n0B8PLUBXwBlYpGv8WEVqkpbEvJJ"
total_cost
0.006432
upstream_id
"msg_01Atpm25yxQVMs1rAb2GeyKp"
created_at
"2024-07-15T09:12:10.106931+00:00"
Monday, July 15th, 2024 at 9:12:10 AM GMT+00:00
model
"anthropic/claude-2.1:beta"
app_id
182717
streamed
false
cancelled
false
provider_name
"Anthropic"
latency
1250
moderation_latency
(null)
generation_time
10
finish_reason
"end_turn"
tokens_prompt
69
tokens_completion
220
native_tokens_prompt
78
native_tokens_completion
242
num_media_prompt
(null)
num_media_completion
(null)
origin
"https://ncbench.com/"
usage
0.006432
Evaluation details
Result
Evaluator
Details
Meta Data
100.0000%
Count dialogue tags
n/a
neededClean
true
dialogueTags
(empty)