INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to names and places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ãĤ©

-0.61

 conflic

-0.56

cause

-0.52

said

-0.51

¿½

-0.50

 nominate

-0.50

 entitlement

-0.49

¶

-0.49

ãĤ¨

-0.48

ä

-0.48

POSITIVE LOGITS

 Heights

0.57

hurst

0.55

igham

0.55

dale

0.53

iac

0.51

 Bros

0.50

 County

0.50

Â®

0.50

ertodd

0.50

witz

0.49

Activations Density 0.612%

proper nouns

proper nouns related to names and places

No Comments

No Known Activations