INDEX

Explanations

vulgar or offensive language

oai_token-act-pair · gpt-3.5-turbo

instances of the word "shit" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

fman

-0.77

HCR

-0.69

Recomm

-0.68

arb

-0.65

 Expend

-0.64

Cour

-0.62

Ctrl

-0.61

Decre

-0.60

 predec

-0.60

tnc

-0.60

POSITIVE LOGITS

bags

1.15

heads

1.11

lords

1.08

storm

1.06

holes

1.03

loads

1.03

fuck

0.99

hole

0.98

wit

0.98

lord

0.98

Activations Density 0.058%

vulgar or offensive language

instances of the word "shit" in various contexts

No Comments

No Known Activations