INDEX

Explanations

references to celebrities

oai_token-act-pair · gpt-3.5-turbo

mentions of celebrities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

doors

-0.71

choes

-0.70

hematic

-0.69

 Wind

-0.69

nerg

-0.66

uv

-0.66

atives

-0.66

empty

-0.65

nda

-0.64

rt

-0.64

POSITIVE LOGITS

rities

1.14

wcs

1.05

 celebrities

1.00

 endors

0.97

 celebrity

0.93

 endorsements

0.89

 celeb

0.85

 gossip

0.81

 superstar

0.78

 Celebrity

0.77

Activations Density 0.012%

references to celebrities

mentions of celebrities

No Comments

No Known Activations