INDEX

Explanations

technical terms and coding-related syntax

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

SequentialGroup

-0.79

RenderAtEndOf

-0.78

 EconPapers

-0.76

AndEndTag

-0.72

 nahilalakip

-0.71

 CreateTagHelper

-0.70

OrNil

-0.68

HtmlAttribute

-0.66

ReusableCell

-0.65

 initComponents

-0.63

POSITIVE LOGITS

 виправивши

0.50

 Infór

0.38

archiviato

0.38

GenerationType

0.37

 EnglishChoose

0.36

 Schulze

0.35

Kesimpulan

0.35

 crossorigin

0.34

providedIn

0.34

 Sobald

0.34

Activations Density 0.006%

technical terms and coding-related syntax

No Comments

No Known Activations