INDEX

Explanations

numerical patterns and references, particularly in scientific contexts

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

SequentialGroup

-0.59

stateMutability

-0.58

NegativeButton

-0.54

-0.52

/***/

-0.52

GOTREF

-0.51

 CallOverrides

-0.50

生平

-0.50

basicConfig

-0.50

exitRule

-0.47

POSITIVE LOGITS

 يتيمه

0.34

numerusform

0.33

әрмәләр

0.33

CrossRef

0.31

dreamstime

0.31

ennemi

0.31

 accidentes

0.30

 enemigo

0.30

 Reisedaten

0.30

werth

0.30

Activations Density 0.184%

numerical patterns and references, particularly in scientific contexts

No Comments

No Known Activations