INDEX

Explanations

statements of fact or reports of actions taken

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

past events and history

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

bootstrapcdn

-0.67

featureID

-0.65

XmlAccessType

-0.64

Personendaten

-0.63

MessageState

-0.63

Vidite

-0.61

Derbyniad

-0.60

Демографія

-0.60

enumii

-0.60

 الحره

-0.60

POSITIVE LOGITS

 history

0.61

 historical

0.53

history

0.51

 HISTORY

0.47

历史

0.46

History

0.45

HISTORY

0.44

 histori

0.44

過去の

0.41

 previous

0.41

Activations Density 0.101%

statements of fact or reports of actions taken

past events and history

No Comments

No Known Activations