INDEX

Explanations

sections related to medical research or treatment outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

ness

-0.35

Ārējās

-0.35

ốm

-0.35

ahan

-0.34

joda

-0.34

mie

-0.34

linien

-0.33

 Beale

-0.32

es

-0.32

dam

-0.31

POSITIVE LOGITS

0.81

Tembelea

0.73

 فريبيس

0.66

AndEndTag

0.59

contentLoaded

0.57

ロウィン

0.57

Tikang

0.56

 zijne

0.56

StoryboardSegue

0.54

 Autorisations

0.54

Activations Density 0.655%

sections related to medical research or treatment outcomes

No Comments

No Known Activations