INDEX

Explanations

technical terms and codes, quantities, and chemical substances

oai_token-act-pair · gpt-3.5-turbo

symbols and characters related to encoded data or formatting

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

hare

-0.62

ettings

-0.56

 strat

-0.54

ortment

-0.53

rontal

-0.52

 Tactics

-0.50

FORE

-0.50

Af

-0.49

forum

-0.49

æĸ¹

-0.49

POSITIVE LOGITS

phia

0.69

ufact

0.59

 herself

0.57

Ã£

0.56

nih

0.56

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.55

iston

0.52

 Sachs

0.50

elson

0.49

chuk

0.47

Activations Density 1.219%

technical terms and codes, quantities, and chemical substances

symbols and characters related to encoded data or formatting

No Comments

No Known Activations