INDEX

Explanations

descriptions of or references to varying degrees of importance or intensity of certain concepts or elements

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a focus on various topics or subjects within discussion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ngth

-0.70

aurus

-0.67

Pwr

-0.66

umers

-0.62

 Thumbnails

-0.62

encia

-0.60

atten

-0.60

lio

-0.60

icone

-0.58

esters

-0.58

POSITIVE LOGITS

 aspects

1.45

 topics

1.28

 aspect

1.15

 facets

1.15

 matters

0.98

 phenomena

0.97

 topic

0.97

 misconceptions

0.97

 particulars

0.95

 facet

0.94

Activations Density 0.399%

descriptions of or references to varying degrees of importance or intensity of certain concepts or elements

phrases indicating a focus on various topics or subjects within discussion

No Comments

No Known Activations