INDEX

Explanations

concepts related to health improvement and wellness

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

 awkwardly

-0.52

الحياه

-0.51

Worse

-0.50

<bos>

-0.48

 underwhelming

-0.48

Etimología

-0.47

やや

-0.47

どうしても

-0.47

degrad

-0.47

 disappointing

-0.46

POSITIVE LOGITS

 healthy

1.22

 healthier

1.17

 happy

1.10

 happier

1.05

Healthy

1.03

 healthiest

1.02

healthy

1.02

 Healthy

1.00

 HAPPY

0.93

 heureuse

0.93

Activations Density 0.208%

concepts related to health improvement and wellness

No Comments

No Known Activations