INDEX

Explanations

words related to manufacturing and making

oai_token-act-pair · gpt-3.5-turbo

words indicating the production or creation of items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

nee

-0.74

 loosen

-0.67

channelAvailability

-0.65

externalActionCode

-0.65

Psy

-0.65

 waive

-0.62

 forfeit

-0.61

---------

-0.61

 oblig

-0.61

 Leone

-0.61

POSITIVE LOGITS

otype

0.88

otypes

0.86

 sure

0.77

 exporting

0.73

ÄŁ

0.73

 batches

0.72

ãĥĥãĤ¯

0.69

itions

0.68

 shipments

0.67

 manufactures

0.65

Activations Density 0.096%

words related to manufacturing and making

words indicating the production or creation of items

No Comments

No Known Activations