INDEX

Explanations

mathematical expressions and symbols in a context that involves logical reasoning or proof structures

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

consecutive punctuation marks

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

 <<<<<<<<<<<<<<

-0.51

 ProtoMessage

-0.46

 honeycomb

-0.44

cetamol

-0.42

 فريبيس

-0.42

 Administrativna

-0.42

ancer

-0.42

GeneratedValue

-0.41

󠁿

-0.41

 chromat

-0.41

POSITIVE LOGITS

 دیکھیے

0.47

 ویکی‌پدی

0.47

@"/

0.46

 AssemblyCompany

0.41

rawDesc

0.38

scryfall

0.38

arakhand

0.37

ValueStyle

0.36

endphp

0.35

SequentialGroup

0.35

Activations Density 0.096%

mathematical expressions and symbols in a context that involves logical reasoning or proof structures

consecutive punctuation marks

No Comments

No Known Activations