INDEX

Explanations

specific named entities and references to importance or relevance in 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

 &___

-0.34

 TextInputType

-0.34

InstrumentedTest

-0.33

łka

-0.31

URIComponent

-0.30

AutoScaleMode

-0.30

 fremden

-0.29

VersionUID

-0.29

 praca

-0.29

 surla

-0.29

POSITIVE LOGITS

 nonUne

0.63

 EconPapers

0.58

 티즈

0.53

 TextAppearance

0.52

strophe

0.52

genie

0.52

fug

0.51

archives

0.51

APL

0.50

 vole

0.50

Activations Density 0.761%

specific named entities and references to importance or relevance in contexts

No Comments

No Known Activations