INDEX

Explanations

specific brand or product names related to various systems and frameworks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tvguidetime

-0.40

 betreft

-0.38

 convenable

-0.36

httphttps

-0.35

Jeografia

-0.31

 egentligen

-0.31

 sicura

-0.30

endaft

-0.30

 derfor

-0.30

ísimas

-0.29

POSITIVE LOGITS

 Argo

0.56

 itſelf

0.54

 ſte

0.53

ArgumentParser

0.53

 paſſ

0.52

ariConfig

0.52

 setUsername

0.52

 Anſ

0.52

ſelf

0.51

leſs

0.50

Activations Density 0.425%

specific brand or product names related to various systems and frameworks

No Comments

No Known Activations