INDEX

Explanations

phrases related to official announcements or statements

oai_token-act-pair · gpt-3.5-turbo

punctuation marks or colons

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

iqueness

-0.67

ilater

-0.66

 recip

-0.66

 upstairs

-0.66

rex

-0.66

sburg

-0.63

 outwe

-0.62

minster

-0.61

angles

-0.61

 carp

-0.60

POSITIVE LOGITS

 Latest

0.94

 Thousands

0.93

 Yesterday

0.92

 Dozens

0.90

 Ahead

0.88

 Following

0.86

 Hundreds

0.85

 Riding

0.83

 Earlier

0.83

 Actor

0.82

Activations Density 0.033%

phrases related to official announcements or statements

punctuation marks or colons

No Comments

No Known Activations