INDEX

Explanations

social media platform mentions

oai_token-act-pair · gpt-3.5-turbo

references to social media platforms, particularly Pinterest and Twitter

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

thro

-0.81

pill

-0.69

iren

-0.68

hid

-0.67

 pill

-0.67

 spoiler

-0.66

©¶æ¥µ

-0.65

liner

-0.63

pled

-0.62

lie

-0.62

POSITIVE LOGITS

 Pinterest

0.85

 PHOTO

0.84

 Photograph

0.84

 Images

0.80

 IMAGES

0.77

atoon

0.74

 Sergeant

0.71

iewicz

0.69

 Painting

0.67

Pic

0.66

Activations Density 0.014%

social media platform mentions

references to social media platforms, particularly Pinterest and Twitter

No Comments

No Known Activations