INDEX

Explanations

phrases expressing personal emotions and opinions

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

程度上

-0.43

理由は

-0.42

 fair

-0.41

shima

-0.41

一応

-0.40

 schien

-0.40

ご理解

-0.40

 Méndez

-0.40

gameserver

-0.39

Proble

-0.39

POSITIVE LOGITS

 '\\;'

1.03

 envy

0.95

 envious

0.92

 adore

0.84

 love

0.84

 LOVE

0.79

LOVE

0.76

 jealous

0.74

 admires

0.74

love

0.73

Activations Density 0.164%

phrases expressing personal emotions and opinions

No Comments

No Known Activations