INDEX

Explanations

references to numeric values or percentages

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 betweenstory

-1.17

transQ

-1.15

 transfieras

-0.98

 Autorizaciones

-0.95

 незавершена

-0.93

 صوتيه

-0.93

تقاوى

-0.92

 queſta

-0.92

 Geſch

-0.91

ьаж

-0.88

POSITIVE LOGITS

0.32

0.31

0.28

0.27

or

0.27

this

0.26

0.25

Activations Density 0.134%

references to numeric values or percentages

No Comments

No Known Activations