INDEX

Explanations

programmatically structured content, particularly template syntax and markup elements

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

 betweenstory

-0.90

OGND

-0.86

EDEFAULT

-0.77

 queſta

-0.76

 BoxFit

-0.74

tagHelperRunner

-0.73

MessageOf

-0.72

HideFlags

-0.67

ſicht

-0.67

 erſt

-0.67

POSITIVE LOGITS

 includes

0.34

 block

0.30

csrf

0.28

<bos>

0.28

 Property

0.27

örté

0.27

 secured

0.27

 preparando

0.27

 entsprechen

0.27

 rappelle

0.26

Activations Density 0.002%

programmatically structured content, particularly template syntax and markup elements

No Comments

No Known Activations