INDEX

Explanations

references to the structure or composition of information

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Савезне

-0.66

parsedMessage

-0.57

 Италијани

-0.47

 Italijani

-0.46

Jereo

-0.45

setVerticalGroup

-0.45

 betweenstory

-0.44

ArrowToggle

-0.44

Jîn

-0.43

LookAnd

-0.42

POSITIVE LOGITS

期刊论文

0.60

tvguidetime

0.54

 surla

0.54

unft

0.50

Viitteet

0.46

 Infórmanos

0.45

rungsseite

0.44

 ujednoznacz

0.43

ophen

0.42

assistant

0.42

Activations Density 0.233%

references to the structure or composition of information

No Comments

No Known Activations