INDEX

Explanations

mentions of actions or activities that involve sharing or consuming items

oai_token-act-pair · gpt-3.5-turbo

references to specific entities or expressions within the text

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

 warr

-0.63

 Siber

-0.61

 sequest

-0.60

 exhib

-0.60

 deduct

-0.59

 Xperia

-0.56

pak

-0.56

 Scand

-0.54

ukong

-0.54

imus

-0.53

POSITIVE LOGITS

aughs

0.71

$.

0.67

ï¸ı

0.67

maxwell

0.67

outh

0.66

emetery

0.65

alian

0.65

thia

0.65

osta

0.64

udder

0.64

Activations Density 0.346%

mentions of actions or activities that involve sharing or consuming items

references to specific entities or expressions within the text

No Comments

No Known Activations