INDEX

Explanations

conjunctions and punctuation marks that indicate lists or connections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 27-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_27/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.27.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

 ſtate

-0.59

 houſe

-0.48

 purpoſe

-0.46

 depositphotos

-0.45

 Majefty

-0.44

 fevere

-0.41

AndEndTag

-0.40

 ويكي

-0.40

 Houſe

-0.40

 preſent

-0.39

POSITIVE LOGITS

rungsseite

0.55

 EconPapers

0.55

 surla

0.54

 utafitiHapana

0.52

ceğine

0.48

IntoConstraints

0.48

istoitu

0.47

IGraphics

0.47

するに

0.46

Rhestr

0.45

Activations Density 0.038%

conjunctions and punctuation marks that indicate lists or connections

No Comments

No Known Activations