INDEX

Explanations

instances of the word "please" indicating requests for action or information

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

please followed by contact or visit

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

findpost

-0.63

sizeCache

-0.54

Datuak

-0.54

DIPSETTING

-0.45

 Мексичка

-0.45

aarrggbb

-0.42

 testigo

-0.42

 määrä

-0.41

 ब्रेकडाउन

-0.40

 utford

-0.40

POSITIVE LOGITS

 visit

0.77

 contact

0.73

contact

0.66

visit

0.63

 click

0.60

 consult

0.59

 Contact

0.56

 contacting

0.55

 reply

0.55

 refer

0.55

Activations Density 0.009%

instances of the word "please" indicating requests for action or information

please followed by contact or visit

No Comments

No Known Activations