INDEX

Explanations

instances of imperatives or commands

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

 linkovi

-0.72

webElementXpaths

-0.70

 <>",

-0.69

دانشنامهٔ

-0.69

Personendaten

-0.67

 يتيمه

-0.61

Personensuche

-0.59

 poveznice

-0.57

 AssemblyTitle

-0.56

contentLoaded

-0.56

POSITIVE LOGITS

 ſte

0.44

BA

0.42

 ſta

0.41

DJ

0.41

 anſ

0.40

 leſs

0.39

Nel

0.39

 Crop

0.38

 faſt

0.38

Look

0.38

Activations Density 0.012%

instances of imperatives or commands

No Comments

No Known Activations