INDEX

Explanations

mentions of educational institutions, particularly those related to technology or law

oai_token-act-pair · gpt-3.5-turbo

instances of the word "of."

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

downs

-0.72

 dots

-0.72

blank

-0.71

NetMessage

-0.70

pmwiki

-0.68

behind

-0.67

aceous

-0.66

acious

-0.65

mere

-0.64

 flush

-0.64

POSITIVE LOGITS

 Excellence

1.03

 Governors

0.79

 Commerce

0.79

 Conference

0.78

 Association

0.77

 Alexandria

0.76

sted

0.75

Commerce

0.75

 Representatives

0.74

 Education

0.73

Activations Density 0.148%

mentions of educational institutions, particularly those related to technology or law

instances of the word "of."

No Comments

No Known Activations