INDEX

Explanations

instances of glimpses or insights into various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

évaluateur

-0.46

 abstrato

-0.45

 Infórmanos

-0.45

 Administrativna

-0.42

 vermelhas

-0.41

harusnya

-0.41

 orgánica

-0.40

 universitaria

-0.40

 bedingt

-0.40

 škole

-0.40

POSITIVE LOGITS

 glimpse

1.41

 glimpses

1.19

 glimp

1.10

 peek

0.82

 insight

0.81

 glance

0.79

 Einblick

0.77

 preview

0.73

peek

0.70

 insights

0.68

Activations Density 0.008%

instances of glimpses or insights into various subjects

No Comments

No Known Activations