INDEX

Explanations

words associated with villains in storytelling

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

sterdam

-0.78

independent

-0.76

DCS

-0.72

galitarian

-0.72

chedel

-0.72

oday

-0.72

ollen

-0.71

JUST

-0.68

independence

-0.65

artney

-0.64

POSITIVE LOGITS

ous

1.40

ously

1.18

 villain

1.01

 Bane

0.99

 villains

0.94

 mastermind

0.91

esses

0.88

ess

0.88

ery

0.81

oid

0.79

Activations Density 0.034%

words associated with villains in storytelling

No Comments

No Known Activations