INDEX

Explanations

statements continuing or elaborating on a previous statement

oai_token-act-pair · gpt-3.5-turbo

phrases indicating explanation or detail

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

ecast

-0.71

 Committees

-0.71

geries

-0.71

wcsstore

-0.70

Born

-0.68

ousands

-0.68

 contracted

-0.66

cko

-0.65

 Donation

-0.64

 Acquisition

-0.64

POSITIVE LOGITS

 sarcast

1.18

 noting

1.05

 omin

1.02

 reiter

1.01

 summar

1.00

 stating

0.99

 elabor

0.97

 emphas

0.97

 pointed

0.95

 quoting

0.94

Activations Density 0.287%

statements continuing or elaborating on a previous statement

phrases indicating explanation or detail

No Comments

No Known Activations