INDEX

Explanations

references to academic or official sources

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

 uVar

-0.52

 surla

-0.50

LAI

-0.41

closedir

-0.40

 مق

-0.39

PRT

-0.39

SSA

-0.39

TDA

-0.37

VOS

-0.36

 AspNetCore

-0.36

POSITIVE LOGITS

 AssemblyTitle

0.63

 wcześ

0.53

 précé

0.50

 earlier

0.50

 previous

0.50

+#+#

0.49

 avoient

0.48

 précédents

0.48

 précédent

0.48

 Inscrivez

0.48

Activations Density 0.354%

references to academic or official sources

No Comments

No Known Activations