INDEX

Explanations

instances of covered or wrapped objects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

ⓧ

-0.63

 متعلقه

-0.62

 sandero

-0.58

 comigo

-0.56

 Doğ

-0.54

 Pliny

-0.54

évaluateur

-0.54

 harps

-0.53

 ayud

-0.51

 quæ

-0.50

POSITIVE LOGITS

gbarkeit

0.71

adpleegd

0.69

*)((

0.67

0.66

NUMX

0.66

Nguồn

0.64

Mutagenicity

0.64

writerow

0.64

 lenker

0.63

phat

0.63

Activations Density 0.329%

instances of covered or wrapped objects

No Comments

No Known Activations