INDEX

Explanations

key terms and phrases related to requests, structure, and definitions in various contexts

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

ⓧ

-0.58

 Италијани

-0.54

 defaultstate

-0.52

jsxFileName

-0.51

 Italijanski

-0.50

arraycopy

-0.48

#+#

-0.47

 CanadaChoose

-0.47

QMetaType

-0.47

uxxxx

-0.46

POSITIVE LOGITS

ᅠ

0.51

PerformLayout

0.50

__(/*!

0.45

dag

0.43

Domin

0.41

ToProps

0.40

گزاری

0.38

гей

0.38

 تعد

0.36

踏み

0.36

Activations Density 0.092%

key terms and phrases related to requests, structure, and definitions in various contexts

No Comments

No Known Activations