INDEX

Explanations

statements about intentions or purposes

oai_token-act-pair · gpt-3.5-turbo

phrases indicating purpose or intention

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

 Towns

-0.69

 knots

-0.65

yes

-0.64

 Watts

-0.60

 Forced

-0.59

 Compass

-0.59

Fra

-0.59

 faults

-0.58

 Years

-0.58

Got

-0.58

POSITIVE LOGITS

 stimulate

1.11

 elicit

1.03

 intimidate

1.02

 discourage

1.02

 emulate

1.02

 dissu

1.00

 provoke

1.00

 alleviate

0.99

 evoke

0.98

 counteract

0.98

Activations Density 0.103%

statements about intentions or purposes

phrases indicating purpose or intention

No Comments

No Known Activations