INDEX

Explanations

references to commercial roles or representations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 שוליים

-0.54

دانشنامهٔ

-0.51

 Dalio

-0.44

 Hospice

-0.42

ครัว

-0.39

tagPool

-0.39

 Flag

-0.39

맘

-0.39

Viungo

-0.39

ınt

-0.38

POSITIVE LOGITS

 celebrity

0.79

 celebrities

0.74

celebrity

0.69

 Celebrity

0.64

 celebs

0.63

Celebrity

0.63

 showbiz

0.60

 entertainers

0.59

 entertainer

0.57

 superstar

0.56

Activations Density 0.227%

references to commercial roles or representations

No Comments

No Known Activations