INDEX

Explanations

proper nouns or names of people or places

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and entities

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

rote

-0.78

actionDate

-0.68

geon

-0.62

redits

-0.61

izons

-0.61

eport

-0.60

@#&

-0.60

wered

-0.58

ledged

-0.58

ãĤ¨ãĥ«

-0.57

POSITIVE LOGITS

 being

1.72

 becoming

1.66

 having

1.59

 needing

1.49

 getting

1.48

 owning

1.45

 gaining

1.40

 discovering

1.40

 losing

1.38

 finding

1.38

Activations Density 0.515%

proper nouns or names of people or places

references to specific individuals and entities

No Comments

No Known Activations