INDEX

Explanations

the word "anch" with varying degrees of similarity

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Panch" and its variations

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

PDATE

-0.68

 Olympics

-0.65

 Jenner

-0.64

Goo

-0.64

DOS

-0.62

berman

-0.59

fts

-0.59

 Scotia

-0.59

pless

-0.59

 FANTASY

-0.58

POSITIVE LOGITS

orage

1.01

arya

0.96

oly

0.95

ovy

0.92

ovies

0.90

aires

0.87

amaru

0.86

aler

0.84

otten

0.83

ieri

0.83

Activations Density 0.008%

the word "anch" with varying degrees of similarity

mentions of the word "Panch" and its variations

No Comments

No Known Activations