INDEX

Explanations

instructions related to using software tools

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

technical instructions

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

 contenido

-0.29

ními

-0.28

 föruts

-0.28

Content

-0.27

getContent

-0.27

Charakter

-0.27

 adult

-0.27

 prison

-0.26

ComponentScan

-0.26

setVerticalGroup

-0.26

POSITIVE LOGITS

 ujednoznacz

0.57

 esternos

0.56

InitVars

0.55

migrationBuilder

0.55

 HasFactory

0.55

featureID

0.55

لينكات

0.55



0.54

AccessorTable

0.54

ühungen

0.54

Activations Density 0.047%

instructions related to using software tools

technical instructions

No Comments

No Known Activations