INDEX

Explanations

phrases related to absence or lack of something

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "nothing."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

assis

-0.72

asio

-0.71

grad

-0.68

CVE

-0.68

NAS

-0.65

srf

-0.63

ilt

-0.63

onduct

-0.62

osen

-0.61

 anonymity

-0.61

POSITIVE LOGITS

 else

1.30

Else

1.13

 whatsoever

0.90

 Else

0.89

 imaginable

0.79

 remotely

0.78

 happens

0.75

 happened

0.73

ness

0.71

 except

0.71

Activations Density 0.037%

phrases related to absence or lack of something

repeated mentions of the word "nothing."

No Comments

No Known Activations