INDEX

Explanations

locations and identifiers related to academic or research affiliations

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

research or academic documentation

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

 Chwiliwch

-0.69

存于互联网档案馆

-0.61

 AssemblyCulture

-0.61

лтемелер

-0.57

Rhestr

-0.57

<unused68>

-0.57

<unused16>

-0.56

<unused52>

-0.56

<unused14>

-0.56

[@BOS@]

-0.56

POSITIVE LOGITS

 doInBackground

0.39

Research

0.39

↵↵

0.39

/*

0.37

package

0.37

/**

0.35

<!--

0.35

Researchers

0.35

Accessors

0.35

 Research

0.34

Activations Density 0.035%

locations and identifiers related to academic or research affiliations

research or academic documentation

No Comments

No Known Activations