INDEX

Explanations

code-related terms specifically linked to user interfaces and touch screen interactions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

 оригіналу

-0.84

 ſtate

-0.82

Efq

-0.76

 Мексичка

-0.75

$_"

-0.74

Datuak

-0.74

RectangleBorder

-0.72

AnimationsModule

-0.71

 Anſ

-0.71

avigation

-0.70

POSITIVE LOGITS

 print

1.26

 printing

1.13

 Print

1.10

print

1.07

Print

1.07

 printer

1.02

 prints

1.02

Printing

1.02

printing

0.98

 printed

0.98

Activations Density 0.221%

code-related terms specifically linked to user interfaces and touch screen interactions

No Comments

No Known Activations