INDEX

Explanations

proper nouns related to the person "Walt."

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Walt."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 tampering

-0.72

 variance

-0.71

 aval

-0.65

 pregn

-0.62

 linkage

-0.61

ysis

-0.61

ulia

-0.60

 elig

-0.60

adesh

-0.60

 saline

-0.60

POSITIVE LOGITS

 Disney

1.08

 Whitman

1.08

rip

0.93

zing

0.91

enthal

0.87

 Walt

0.87

stad

0.83

terson

0.79

 Meier

0.79

zman

0.79

Activations Density 0.014%

proper nouns related to the person "Walt."

mentions of the name "Walt."

No Comments

No Known Activations