INDEX

Explanations

statements and phrases that suggest implications or associations related to social or cultural beliefs

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

phrases expressing implications or suggesting something without directly stating it.

oai_token-act-pair · claude-3-5-haiku-20241022 Triggered by @neilrathi

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Noice

-0.46

 الحره

-0.45

farwyddwr

-0.45

notices

-0.44

 unle

-0.43

pherd

-0.42

 vermogen

-0.42

 Asympto

-0.42

InputBorder

-0.42

を開催

-0.41

POSITIVE LOGITS

 implying

0.46

enumii

0.45

RTDA

0.43

 implied

0.40

OCCURRED

0.39

 quelconque

0.38

 imply

0.38

 implication

0.36

المكان

0.36

 insinu

0.35

Activations Density 0.218%

statements and phrases that suggest implications or associations related to social or cultural beliefs

phrases expressing implications or suggesting something without directly stating it.

No Comments

No Known Activations