INDEX

Explanations

various forms of visual or textual formatting and engagement indicators

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 мә

-0.38

fø

-0.38

 пу

-0.37

thy

-0.37

 propria

-0.34

hour

-0.34

UALA

-0.34

 muun

-0.34

أن

-0.34

-0.33

POSITIVE LOGITS

Personendaten

0.96

 kasarigan

0.93

帖最后由

0.91

titleMargin

0.86

addCriterion

0.84

MigrationBuilder

0.84

enterOuterAlt

0.84

jsxFileName

0.82

twimg

0.80

 myſelf

0.80

Activations Density 0.486%

various forms of visual or textual formatting and engagement indicators

No Comments

No Known Activations