INDEX

Explanations

numerical identifiers and error codes

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

cshtml

-0.53

 bowiem

-0.52

mobileqq

-0.50

arbox

-0.49

ousine

-0.49

 initComponents

-0.49

halve

-0.48

uete

-0.48

原始内容存档于

-0.47

dersfield

-0.47

POSITIVE LOGITS

principalColumn

0.34

ibatkan

0.34

IUrlHelper

0.34

Décès

0.34

.*")]

0.31

لينكات

0.30

✭✭

0.30

Indicates

0.29

 marbre

0.29

ReusableCell

0.29

Activations Density 0.060%

numerical identifiers and error codes

No Comments

No Known Activations