INDEX

Explanations

contrasting pairs of words and phrases

oai_token-act-pair · gpt-3.5-turbo

words and phrases indicating comparisons or contrasts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

iband

-0.76

ONSORED

-0.76

ktop

-0.72

utenberg

-0.72

nown

-0.68

auder

-0.68

oren

-0.66

ITED

-0.66

 "$:/

-0.66

akespe

-0.65

POSITIVE LOGITS

 Magic

0.68

 fashion

0.63

 booze

0.63

Magic

0.62

 messenger

0.61

 sorcery

0.61

 Dove

0.60

 rabbi

0.60

ãĥīãĥ©

0.60

CEO

0.60

Activations Density 0.660%

contrasting pairs of words and phrases

words and phrases indicating comparisons or contrasts

No Comments

No Known Activations