INDEX

Explanations

proper nouns and specific entities, such as names, organizations, and locations

oai_token-act-pair · gpt-3.5-turbo

terms and phrases related to commonality and rarity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ilee

-0.57

elight

-0.54

odox

-0.53

hemer

-0.51

emonium

-0.51

vertisement

-0.51

ivably

-0.51

inventoryQuantity

-0.50

new

-0.50

kered

-0.49

POSITIVE LOGITS

{"

0.40

 Snapdragon

0.39

$$

0.39

.","

0.37

"},"

0.37

MD

0.37

 Basketball

0.36

AR

0.36

CTR

0.36

``

0.36

Activations Density 1.757%

proper nouns and specific entities, such as names, organizations, and locations

terms and phrases related to commonality and rarity

No Comments

No Known Activations