INDEX

Explanations

elements related to narratives of personal experiences or interactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

 périodes

-0.44

 vermogen

-0.43

 sogget

-0.43

 nadzieję

-0.42

 Wappen

-0.41

uș

-0.40

 umane

-0.40

 praktijk

-0.40

بيرة

-0.40

ślę

-0.39

POSITIVE LOGITS

 someone

1.21

someone

1.15

Someone

1.09

 somebody

1.07

 Someone

1.05

somebody

0.99

 thieves

0.94

Somebody

0.93

 hordes

0.91

 SOMEONE

0.89

Activations Density 0.392%

elements related to narratives of personal experiences or interactions

No Comments

No Known Activations