INDEX

Explanations

phrases related to daily news updates and recommendations

oai_token-act-pair · gpt-3.5-turbo

mentions of curated selections or recommendations

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

lia

-0.67

gal

-0.63

amen

-0.59

reb

-0.59

 nerv

-0.56

gren

-0.56

 territ

-0.56

bum

-0.56

 peac

-0.56

 pige

-0.56

POSITIVE LOGITS

 Picks

0.90

ļéĨĴ

0.67

UGC

0.65

ettings

0.65

ilver

0.63

heet

0.62

eter

0.62

 Secrets

0.61

ãĤ¨ãĥ«

0.61

leep

0.61

Activations Density 0.023%

phrases related to daily news updates and recommendations

mentions of curated selections or recommendations

No Comments

No Known Activations