INDEX

Explanations

instances of the word "even" followed by other content

oai_token-act-pair · gpt-3.5-turbo

instances of the word "even" to indicate escalation or inclusivity in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

aim

-0.85

ffen

-0.84

ierrez

-0.82

isen

-0.73

rend

-0.72

cent

-0.71

aliation

-0.70

Ĥİ

-0.70

unity

-0.69

ayers

-0.68

POSITIVE LOGITS

 worse

0.93

 downright

0.86

handedly

0.85

 stranger

0.84

 outright

0.83

 though

0.82

 remotely

0.78

 nicer

0.67

 thrive

0.65

 paran

0.65

Activations Density 0.050%

instances of the word "even" followed by other content

instances of the word "even" to indicate escalation or inclusivity in various contexts

No Comments

No Known Activations