INDEX

Explanations

mentions of the word "ard" at different activation levels

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of "guardianship" or "protection."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

PER

-0.93

pheus

-0.84

ĸļ

-0.80

ĺħ

-0.77

igslist

-0.76

raltar

-0.74

avascript

-0.68

 Pastebin

-0.66

phis

-0.66

uyomi

-0.65

POSITIVE LOGITS

ragon

1.00

robe

0.93

ice

0.91

rill

0.91

furt

0.85

 Mechdragon

0.85

iop

0.84

rums

0.82

aign

0.80

uce

0.79

Activations Density 0.049%

mentions of the word "ard" at different activation levels

terms related to the concept of "guardianship" or "protection."

No Comments

No Known Activations