INDEX

Explanations

references to user roles and group classifications within a structured data format

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

user roles and groups

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

AndEndTag

-0.39

 scra

-0.37

unse

-0.35

hot

-0.34

clusal

-0.34

 combinação

-0.34

hot

-0.33

 catchy

-0.32

 packing

-0.30

thol

-0.30

POSITIVE LOGITS

 userType

0.69

GrantedAuthority

0.67

MigrationBuilder

0.66

NameInMap

0.62

 roles

0.59

roles

0.55

 Roles

0.53

getRole

0.52

 Италијани

0.52

userType

0.52

Activations Density 0.004%

references to user roles and group classifications within a structured data format

user roles and groups

No Comments

No Known Activations