INDEX

Explanations

names and identifiers in technical or structured contexts

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

expandindo

-0.51

RegressionTest

-0.50

 ujednoznacz

-0.47

saraba

-0.46

 FontWeight

-0.46

 Lokales

-0.44

 vulcan

-0.41

 lordship

-0.40

DIPSETTING

-0.39

 Lordship

-0.39

POSITIVE LOGITS

AndEndTag

0.54

ValueStyle

0.42

/*

0.38

Parcelize

0.37

Билгалдахарш

0.36

AccessorTable

0.36

 tetto

0.36

 humanidade

0.35

 الرياضيه

0.34

habitude

0.34

Activations Density 1.897%

names and identifiers in technical or structured contexts

No Comments

No Known Activations