INDEX

Explanations

the frequency of the word "done" and its variations in different contexts

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

completion of actions

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

SourceChecksum

-0.56

ViewFeatures

-0.49

DataAnnotations

-0.47

__(/*!

-0.46

 Exposed

-0.41

 již

-0.40

})}\

-0.40

 UIApplication

-0.40

Datuak

-0.39

LMS

-0.39

POSITIVE LOGITS

 برانيه

0.58

 crossorigin

0.58

 FetchType

0.58

 pegat

0.57

 enggak

0.54

DIPSETTING

0.52

 bigger

0.52

 cewek

0.50

 Anſ

0.50

 pelea

0.49

Activations Density 0.423%

the frequency of the word "done" and its variations in different contexts

completion of actions

No Comments

No Known Activations