INDEX

Explanations

persistent usage of symbols and formatting in digital contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.17.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

 unknownFields

-0.58

+#+#

-0.48

aarrggbb

-0.47

 nakalista

-0.46

shimano

-0.42

Linki

-0.42

Hozzáférés

-0.42

ntesis

-0.41

 OMITTED

-0.41

ConstraintMaker

-0.41

POSITIVE LOGITS

tvguidetime

0.78

pholes

0.69

 endIndex

0.62

 JsonConvert

0.60

ykjav

0.60

 tartalomajánló

0.58

rinas

0.58

KURZBESCHREIBUNG

0.58

setItem

0.56

rubin

0.56

Activations Density 0.026%

persistent usage of symbols and formatting in digital contexts

No Comments

No Known Activations