INDEX

Explanations

phrases indicating affirmation or confirmation

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

weibo

-0.44

Communist

-0.43

時から

-0.43

 동안

-0.43

 eleg

-0.41

//});

-0.39

하세요

-0.39

 Robb

-0.39

 üzere

-0.39

PCD

-0.39

POSITIVE LOGITS

 zijne

0.61

 zelve

0.60

 faptul

0.60

PreferredItem

0.60

 kasarigan

0.59

 insuffisamment

0.58

 zoude

0.57

 dezelve

0.56

 fapt

0.56

erintah

0.56

Activations Density 0.440%

phrases indicating affirmation or confirmation

No Comments

No Known Activations