INDEX

Explanations

instances of written communication or authorship

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

complexContent

-0.68

 سكانية

-0.65

RTEX

-0.62

WebElementEntity

-0.61

 myſelf

-0.58

 pleaſure

-0.58

醐

-0.57

 ddelweddau

-0.56

ieR

-0.56

 itſelf

-0.56

POSITIVE LOGITS

 wrote

0.88

 write

0.81

write

0.79

 écri

0.70

 Write

0.70

Write

0.67

 writing

0.67

 writes

0.67

 escribió

0.66

 menulis

0.65

Activations Density 0.007%

instances of written communication or authorship

No Comments

No Known Activations