INDEX

Explanations

phrases that indicate context or conditions surrounding an action or situation

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

in part because

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

AutoField

-0.64

󠁮

-0.57

 noDo

-0.55

 Wiktionnaire

-0.53

kháu

-0.51

matchCondition

-0.51

getMinutes

-0.50

Erstellt

-0.49

énéral

-0.49

 uitz

-0.49

POSITIVE LOGITS

Parcelize

0.40

ことも

0.39

 partly

0.35

besondere

0.33

 "..\..\..\

0.33

 Helped

0.33

 encourages

0.33

webElementXpaths

0.33

tamment

0.32

dont

0.32

Activations Density 0.055%

phrases that indicate context or conditions surrounding an action or situation

in part because

No Comments

No Known Activations