INDEX

Explanations

references to choices and specific abilities related to personal or technological empowerment

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

"The" followed by a word indicating possibility or capability

np_acts-logits-general · gemini-2.0-flash

potential or ability

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 forState

-0.43

いかに

-0.42

gms

-0.39

पया

-0.39

feof

-0.36

 denen

-0.36

spesies

-0.35

AndEndTag

-0.35

kir

-0.35

zadas

-0.34

POSITIVE LOGITS

 ability

0.71

 opportunity

0.65

 chance

0.59

 necessary

0.59

 privilege

0.59

 capacity

0.58

 capability

0.57

 option

0.56

ScopeManager

0.54

 means

0.53

Activations Density 0.126%

references to choices and specific abilities related to personal or technological empowerment

"The" followed by a word indicating possibility or capability

potential or ability

No Comments

No Known Activations