INDEX

Explanations

phrases and terminology related to grammatical structures and verb conjugations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.38.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

 enfans

-1.31

 nôtre

-1.22

 plumas

-1.20

 Füßen

-1.20

 orejas

-1.18

 Wikiseite

-1.18

 Menschheit

-1.17

 Decke

-1.16

 Geruch

-1.16

 Männern

-1.15

POSITIVE LOGITS

mit

0.68

neu

0.63

die

0.60

ver

0.59

sch

0.58

bis

0.57

von

0.56

 dyna

0.56

ent

0.56

und

0.56

Activations Density 0.237%

phrases and terminology related to grammatical structures and verb conjugations

No Comments

No Known Activations