INDEX

Explanations

structured data in the form of lists or options

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

div elements

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

 pinulongan

-0.63

ViewFeatures

-0.59

jspb

-0.56

الحياه

-0.54

🔕

-0.53

DebuggerNonUser

-0.51

󠁬

-0.49

⟬

-0.48

 ſta

-0.47

mese

-0.47

POSITIVE LOGITS

 eksemp

0.33

fél

0.33

endforeach

0.32

]=>

0.32

 betweenstory

0.32

 mérit

0.32

 mérite

0.31

 vertes

0.31

 createSlice

0.31

rungsseite

0.29

Activations Density 0.054%

structured data in the form of lists or options

div elements

No Comments

No Known Activations