INDEX

Explanations

phrases related to arrivals and returns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 préféré

-0.43

 privilégi

-0.41

 jezelf

-0.36

hubung

-0.35

eachtet

-0.35

verze

-0.35

ấc

-0.35

 lunares

-0.34

 snippetHide

-0.34

ciclop

-0.33

POSITIVE LOGITS

 arrival

0.82

InputBorder

0.77

 Arrival

0.77

arrival

0.74

 arrivals

0.74

 arrive

0.73

Arrival

0.72

 arrives

0.71

Arrive

0.69

 Arrive

0.69

Activations Density 0.450%

phrases related to arrivals and returns

No Comments

No Known Activations