INDEX

Explanations

phrases indicating uncertainty or conditions of evidence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 however

-0.56

 però

-0.48

however

-0.43

 entanto

-0.42

However

-0.41

ernoo

-0.41

RenderAtEndOf

-0.40

 However

-0.39

 الاطلاع

-0.39

reportWebVitals

-0.38

POSITIVE LOGITS

 ohnehin

0.57

ArrowToggle

0.57

يكب

0.52

Personensuche

0.48

 &___

0.48

そもそも

0.45

 avoient

0.45

 sowieso

0.45

principalTable

0.43

何より

0.42

Activations Density 0.724%

phrases indicating uncertainty or conditions of evidence

No Comments

No Known Activations