INDEX

Explanations

non-English characters or coding symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

########.

-0.82

 مشين

-0.80

 ModelExpression

-0.77

 للمعارف

-0.69

 متعلقه

-0.66

 initComponents

-0.66

 OnInit

-0.65

.*")]

-0.61

+#+#

-0.60

UnusedPrivate

-0.60

POSITIVE LOGITS

 délib

0.56

<",

0.53

 whor

0.52

madu

0.51

 pinn

0.50

colours

0.49

 parson

0.48

 Emin

0.48

 Colours

0.48

 shroud

0.48

Activations Density 0.151%

non-English characters or coding symbols

No Comments

No Known Activations