INDEX

Explanations

elements related to years and metadata labels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 25-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_25/width_16k/average_l0_116

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.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

 ſtate

-0.55

 purpoſe

-0.53

 houſe

-0.50

 Ortsteil

-0.50

 perpetuity

-0.49

 myſelf

-0.48

 Shaksp

-0.47

 Esau

-0.46

 reaſon

-0.46

 behalf

-0.46

POSITIVE LOGITS

 maken

0.57

 komen

0.57

 gaan

0.56

 nemen

0.56

 doen

0.55

ubern

0.54

 brengen

0.54

 leggen

0.54

poken

0.54

 houden

0.53

Activations Density 0.013%

elements related to years and metadata labels

No Comments

No Known Activations