INDEX

Explanations

references to locations and entities, particularly in legal and governmental 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-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

 pinulongan

-0.69

batore

-0.65

SharedCtor

-0.64

 resourceCulture

-0.62

\{\\

-0.61

 utafitiHapana

-0.60

__(/*!

-0.60

msgTypes

-0.58

Personendaten

-0.57

 wireType

-0.54

POSITIVE LOGITS

 Potomac

0.78

 Washington

0.78

 Canberra

0.74

Washington

0.73

WASHINGTON

0.68

 washington

0.65

TAWA

0.64

washingtonpost

0.63

 WASHINGTON

0.60

 Delhi

0.60

Activations Density 0.608%

references to locations and entities, particularly in legal and governmental contexts

No Comments

No Known Activations