INDEX

Explanations

terms related to legal or medical contexts, particularly around incarceration and conditions of parole

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 propOrder

-0.63

 ring

-0.59

fly

-0.57

 land

-0.56

 water

-0.55

 book

-0.54

 wall

-0.53

 work

-0.53

 house

-0.53

 heat

-0.52

POSITIVE LOGITS

verwijspagina

0.62

ſelf

0.59

kuuta

0.55

 betweenstory

0.54

 Theſe

0.54

ſelves

0.53

Efq

0.53

ciaż

0.52

meadow

0.51

tvguidetime

0.50

Activations Density 0.885%

terms related to legal or medical contexts, particularly around incarceration and conditions of parole

No Comments

No Known Activations