INDEX

Explanations

titles or mentions of authors

oai_token-act-pair · gpt-3.5-turbo

the word "author" and its variations

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

ll

-0.75

 Libre

-0.71

tone

-0.67

EMS

-0.67

 Nicaragua

-0.66

fw

-0.65

 Eastern

-0.63

 Territories

-0.63

 Bots

-0.62

ijn

-0.62

POSITIVE LOGITS

itatively

1.61

itative

1.16

itar

1.13

itarian

1.10

essee

0.95

itent

0.94

itism

0.87

izations

0.87

ournals

0.86

isations

0.85

Activations Density 0.031%

titles or mentions of authors

the word "author" and its variations

No Comments

No Known Activations