INDEX

Explanations

the letter 'e' in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Metzger

-0.48

 này

-0.42

 của

-0.42

 Wartawan

-0.40

Außerdem

-0.40

########.

-0.39

<eos>

-0.39

 untuk

-0.38

 初始化

-0.38

 ของ

-0.38

POSITIVE LOGITS

tvguidetime

0.69

 ainfi

0.67

GEBURTS

0.66

tonode

0.65

 propOrder

0.63

 kasarigan

0.62

LookAnd

0.61

 quæ

0.61

<bos>

0.60

 auroit

0.59

Activations Density 0.033%

the letter 'e' in various contexts

No Comments

No Known Activations