INDEX

Explanations

terms related to text messaging and communication etiquette

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

 Anlass

-0.39

 Finanzierung

-0.37

PhysRevD

-0.36

 situés

-0.36

 ubicada

-0.35

 Erhaltung

-0.35

 Italijanski

-0.35

 Schwerpunkt

-0.35

 Manusia

-0.35

 Auffassung

-0.34

POSITIVE LOGITS

 texting

0.86

 texted

0.75

sms

0.72

 messages

0.70

 snippetHide

0.69

makeText

0.67

 telegram

0.66

SMS

0.64

MessageState

0.64

 texts

0.63

Activations Density 0.011%

terms related to text messaging and communication etiquette

No Comments

No Known Activations