INDEX

Explanations

references to specific individuals or characters

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

 saba

-0.42

 cheap

-0.42

cheap

-0.40

搭

-0.40

arrow

-0.40

 Sandro

-0.40

 arrow

-0.40

Dougall

-0.39

ostringstream

-0.38

壕

-0.38

POSITIVE LOGITS

LG

0.63

ele

0.63

rungsseite

0.59

LG

0.57

lg

0.53

ier

0.52

ele

0.52

 zijne

0.51

 newOwner

0.51

rlpool

0.50

Activations Density 0.365%

references to specific individuals or characters

No Comments

No Known Activations