INDEX

Explanations

elements related to measurements or counts

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pal

-0.40

vig

-0.38

 DOMIN

-0.37

rentiel

-0.34

 grand

-0.34

Pal

-0.34

Fais

-0.34

inien

-0.34

 Divine

-0.33

Divine

-0.32

POSITIVE LOGITS

AddHtmlAttribute

0.66

hoeddwyd

0.66

Diweddarwch

0.60

hyrchwyd

0.52



0.51

 autorytatywna

0.47

 ProtoMessage

0.46

PyTuple

0.46

Autoritní

0.46

webElementXpaths

0.46

Activations Density 0.995%

elements related to measurements or counts

No Comments

No Known Activations