INDEX

Explanations

phrases indicating a recent accomplishment or activity

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a person's recent experiences or status changes

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

kHz

-0.76

ilation

-0.75

%]

-0.75

kHz

-0.72

Hz

-0.72

%:

-0.71

MSN

-0.70

Enlarge

-0.68

ãĤ±

-0.67

å¥³

-0.66

POSITIVE LOGITS

 defeats

0.77

 defeat

0.76

 victories

0.72

bye

0.68

eatures

0.67

wre

0.64

 leash

0.63

 pist

0.63

stage

0.63

ibaba

0.62

Activations Density 0.019%

phrases indicating a recent accomplishment or activity

phrases indicating a person's recent experiences or status changes

No Comments

No Known Activations