INDEX

Explanations

phrases related to product or item descriptions

oai_token-act-pair · gpt-3.5-turbo

phrases related to memories or significant experiences

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

paio

-0.77

Recomm

-0.74

icter

-0.73

opian

-0.70

 sovere

-0.70

TPPStreamerBot

-0.69

edIn

-0.68

Secondly

-0.68

Ire

-0.67

llah

-0.66

POSITIVE LOGITS

 FANTASY

0.75

 Cancel

0.71

ategories

0.69

hester

0.69

MEN

0.68

 Selected

0.68

inav

0.67

ategor

0.67

 Beginning

0.66

 Vampire

0.66

Activations Density 3.378%

phrases related to product or item descriptions

phrases related to memories or significant experiences

No Comments

No Known Activations