INDEX

Explanations

names or titles related to various entities or characters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

WithIOException

-0.62

 Inſ

-0.60

duled

-0.55

 poffible

-0.55

ſelf

-0.54

parsedMessage

-0.54

 témoig

-0.53

 ſte

-0.52

 deſſen

-0.52

licitations

-0.51

POSITIVE LOGITS

 surla

0.45

hyrchwyd

0.39

 lèvres

0.38

iecie

0.36

Espèce

0.36

それだけ

0.34

 vivere

0.33

 Wikispecies

0.33

Дереккөздер

0.33

ViewImports

0.33

Activations Density 0.216%

names or titles related to various entities or characters

No Comments

No Known Activations