INDEX

Explanations

phrases and variations related to the action of wrapping or enclosing items

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

wrapping and wrapped objects

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

DatabaseError

-0.59

Sain

-0.54

contentLoaded

-0.52

醐

-0.46

oy

-0.45

addCriterion

-0.45

UIControlState

-0.43

ien

-0.43

huy

-0.42

";"

-0.42

POSITIVE LOGITS

 Wrap

1.18

wrap

1.14

Wrap

1.12

 wrap

1.11

WRAP

1.09

 WRAP

1.08

 Wrapping

1.06

 wrapping

1.04

 Wraps

0.96

Wrapping

0.96

Activations Density 0.008%

phrases and variations related to the action of wrapping or enclosing items

wrapping and wrapped objects

No Comments

No Known Activations