INDEX

Explanations

names, references to individuals, and titles

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

ending in "rol," "rel," "hel,""el," or "al"

np_acts-logits-general · gemini-2.0-flash

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

ように

-0.92

으로

-0.80

less

-0.64

länge

-0.64

ٔ

-0.62

ності

-0.57

ا

-0.57

 noastră

-0.54

ities

-0.52

 으로

-0.52

POSITIVE LOGITS

led

0.69

WriteTagHelper

0.57

 ExecuteAsync

0.54

ScopeManager

0.52

lel

0.52

llllllll

0.49

al

0.48

ateral

0.46

pnl

0.46

tal

0.46

Activations Density 1.306%

names, references to individuals, and titles

ending in "rol," "rel," "hel,""el," or "al"

No Comments

No Known Activations