INDEX

Explanations

phrases related to specific demographics or social identities

oai_token-act-pair · gpt-3.5-turbo

phrases indicating attribution or sources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 derail

-0.67

escal

-0.65

 waste

-0.63

lessly

-0.62

 backlog

-0.61

RG

-0.61

 fruitful

-0.60

 salv

-0.60

icular

-0.59

 shel

-0.59

POSITIVE LOGITS

 Mechdragon

0.69

retty

0.68

ð

0.66

 Apprentice

0.66

inner

0.66

 Creed

0.66

éŃĶ

0.65

 Percentage

0.64

terday

0.64

WT

0.64

Activations Density 0.503%

phrases related to specific demographics or social identities

phrases indicating attribution or sources

No Comments

No Known Activations