INDEX

Explanations

references to conspiracy theories and related terminology

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

conspiracy theories, Deep State

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

 Intermediate

-0.35

เฉ

-0.35

sam

-0.34

inflater

-0.34

 getItemId

-0.33

 Caring

-0.33

MEDIATE

-0.33

 intermediate

-0.32

مر

-0.31

 immediate

-0.31

POSITIVE LOGITS

CppCodeGen

0.71

tvguidetime

0.67

BufferException

0.65

 estekak

0.64

PickerController

0.62

 nahilalakip

0.59

 ujednoznacz

0.59

AddWithValue

0.57

 مشين

0.56

 Obrador

0.56

Activations Density 0.090%

references to conspiracy theories and related terminology

conspiracy theories, Deep State

No Comments

No Known Activations