INDEX

Explanations

action and event-related words

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

following quantifiers or comparatives

np_acts-logits-general · gemini-2.0-flash

allwhocommitsvocalthe

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

spinal

-0.56

 فريبيس

-0.56

spun

-0.52

LayoutConstraint

-0.49

spinning

-0.49

SourceChecksum

-0.49

 BoxFit

-0.49

fron

-0.48

spin

-0.48

MetaObject

-0.47

POSITIVE LOGITS

 menn

0.37

ambién

0.35

ピール

0.34

DeleteMapping

0.33

envolvimento

0.32

CppCodeGen

0.32

 Inscrivez

0.32

paravant

0.31

:✨

0.31

 voices

0.30

Activations Density 0.194%

action and event-related words

following quantifiers or comparatives

allwhocommitsvocalthe

No Comments

No Known Activations