INDEX

Explanations

quantifiers that indicate a large number or variety

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

quantifiers followed by nouns

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

awtextra

-0.58

Personendaten

-0.57

IUrlHelper

-0.56

 ujednoznacz

-0.49

 Numerade

-0.48

EndContext

-0.48

VIAF

-0.47

setVerticalGroup

-0.47

 GenerationType

-0.45

MemoryWarning

-0.44

POSITIVE LOGITS

 various

0.55

 relevant

0.54

 other

0.51

 prominent

0.51

 bright

0.50

 popular

0.50

 numerous

0.49

 brilliant

0.48

 positive

0.48

 pure

0.48

Activations Density 0.038%

quantifiers that indicate a large number or variety

quantifiers followed by nouns

No Comments

No Known Activations