INDEX

Explanations

the word "guess" followed by other words or phrases

oai_token-act-pair · gpt-3.5-turbo

instances of the expression "guess what" or similar phrases indicating anticipation or rhetorical questions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iences

-0.84

interrupted

-0.75

contained

-0.74

ĸļ

-0.70

Interstitial

-0.66

enture

-0.65

acca

-0.64

andals

-0.63

agos

-0.63

EVA

-0.62

POSITIVE LOGITS

 guesses

1.09

 guess

0.99

 guessing

0.88

 Guess

0.79

itial

0.77

 guessed

0.77

 incorrectly

0.71

work

0.70

 darn

0.69

yp

0.69

Activations Density 0.043%

the word "guess" followed by other words or phrases

instances of the expression "guess what" or similar phrases indicating anticipation or rhetorical questions

No Comments

No Known Activations