INDEX

Explanations

numerical values and character identifiers related to classifications or categories

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

classification codes

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

WebElementEntity

-0.61

WriteTagHelper

-0.58

 surla

-0.55

 kasarigan

-0.54

ագրություններ

-0.53

mybatisplus

-0.47

exitRule

-0.45

Personendaten

-0.42

MessageTagHelper

-0.42

солю

-0.42

POSITIVE LOGITS

HostException

0.48

 arada

0.45

 tampoco

0.44

それに

0.44

ãy

0.43

 zubehör

0.42

 večer

0.42

 ebenfalls

0.41

 sekitarnya

0.41

 habet

0.41

Activations Density 0.006%

numerical values and character identifiers related to classifications or categories

classification codes

No Comments

No Known Activations