INDEX

Explanations

terms related to specific concepts, such as body image issues and military units

oai_token-act-pair · gpt-3.5-turbo

names or references associated with specific individuals, places, or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

rez

-0.65

CB

-0.60

ERG

-0.57

US

-0.57

 Walls

-0.57

bern

-0.56

 Carrier

-0.56

bear

-0.54

RIS

-0.54

 STATS

-0.54

POSITIVE LOGITS

ioxide

0.98

merce

0.72

amily

0.68

enei

0.67

 imitation

0.67

 necklace

0.67

estyles

0.65

 charact

0.65

 dime

0.64

nesium

0.63

Activations Density 0.374%

terms related to specific concepts, such as body image issues and military units

names or references associated with specific individuals, places, or events

No Comments

No Known Activations