INDEX

Explanations

words related to technology companies and products

oai_token-act-pair · gpt-3.5-turbo

possessive forms and words associated with brands or products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Reviewer

-0.94

Filename

-0.94

requisites

-0.83

Appearances

-0.77

obin

-0.75

iece

-0.75

Posts

-0.75

sectional

-0.74

Anonymous

-0.73

Redditor

-0.73

POSITIVE LOGITS

own

1.10

 flagship

1.10

 proprietary

1.07

 newest

1.03

 foray

1.00

 patented

0.98

 forthcoming

0.95

 upcoming

0.92

 venerable

0.89

 latest

0.89

Activations Density 0.186%

words related to technology companies and products

possessive forms and words associated with brands or products

No Comments

No Known Activations