INDEX

Explanations

code-related phrases or expressions, particularly those involving functions and output statements

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

 يتيمه

-0.52

 <>",

-0.42

balleur

-0.39

RectangleBorder

-0.38

ftagPool

-0.36

QMetaType

-0.36

المناصب

-0.36

ToBounds

-0.36

abestanden

-0.35

цездатний

-0.34

POSITIVE LOGITS

gdx

0.60

JspWriter

0.57

xrLabel

0.56

 text

0.48

writeln

0.46

sheet

0.46

Text

0.46

paj

0.45

CommandText

0.44

WidgetItem

0.44

Activations Density 0.316%

code-related phrases or expressions, particularly those involving functions and output statements

No Comments

No Known Activations