INDEX

Explanations

words related to negative or problematic situations and actions

oai_token-act-pair · gpt-3.5-turbo

terms related to disruption or disservice

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Juliet

-0.70

 Suit

-0.67

 Goldberg

-0.62

 Shogun

-0.62

 drafts

-0.62

 Goose

-0.61

OPLE

-0.61

Eve

-0.60

DragonMagazine

-0.59

 Isabel

-0.59

POSITIVE LOGITS

rup

1.25

array

1.24

assemb

1.21

concert

1.19

organized

1.17

qus

1.16

quiet

1.16

burse

1.15

interested

1.14

ordered

1.13

Activations Density 0.014%

words related to negative or problematic situations and actions

terms related to disruption or disservice

No Comments

No Known Activations