INDEX

Explanations

specific web addresses or domain names

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

domain names

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

ScopeManager

-0.73

//

-0.62

IsPostBack

-0.57

AddTagHelper

-0.54

Personendaten

-0.52

 raiſ

-0.48

:✨

-0.48

 ModelExpression

-0.48

 CreateTagHelper

-0.47

 للاسماء

-0.47

POSITIVE LOGITS

ษัท

0.44

 Савезне

0.43

calça

0.42

จัก

0.39

ramienta

0.39

喱

0.39

Jeografia

0.39

 AssemblyCompany

0.37

 kasar

0.36

 Pferd

0.36

Activations Density 0.020%

specific web addresses or domain names

domain names

No Comments

No Known Activations