INDEX

Explanations

repeated phrases and references to organizations and programs

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

IntoConstraints

-0.64

MethodManager

-0.54

 configureStore

-0.49

 JpaRepository

-0.47

 giù

-0.47

runApp

-0.47

MemoryWarning

-0.46

 mijne

-0.46

VYMaps

-0.45

DIPSETTING

-0.45

POSITIVE LOGITS

 ausp

0.44

 outreach

0.44

 Programme

0.42

AddTo

0.42

 programmes

0.42

 programme

0.41

cuadro

0.41

 Program

0.41

 Perlu

0.41

 Programmes

0.40

Activations Density 0.025%

repeated phrases and references to organizations and programs

No Comments

No Known Activations