INDEX

Explanations

programming-related terms and structures

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 مرئيه

-0.91

 MainAxisSize

-0.85

AddHtmlAttribute

-0.82

IBOutlet

-0.76

Tikang

-0.73

 IBOutlet

-0.73

ReusableCell

-0.73

AddTagHelper

-0.72

Rohy

-0.72

 فريبيس

-0.71

POSITIVE LOGITS

 Constitucional

0.41

 Moscú

0.39

 Irán

0.37

behandling

0.36

midler

0.35

 katanya

0.34

 Turquía

0.33

 Meksika

0.33

 Fassung

0.33

Activations Density 0.352%

programming-related terms and structures

No Comments

No Known Activations