INDEX

Explanations

Instagram-related terms

oai_token-act-pair · gpt-3.5-turbo

references to the social media platform Instagram

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

inelli

-0.66

 Booker

-0.66

 Norwich

-0.65

 serious

-0.64

Beg

-0.64

 Isles

-0.62

Tob

-0.62

CES

-0.62

 Carson

-0.61

ìĿ

-0.61

POSITIVE LOGITS

agram

1.63

matical

1.12

matically

0.92

yrinth

0.90

uration

0.89

matic

0.89

urated

0.89

tags

0.88

ombies

0.82

alogy

0.81

Activations Density 0.008%

Instagram-related terms

references to the social media platform Instagram

No Comments

No Known Activations