INDEX

Explanations

instances of dialogue or reported speech

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

words related to speech and quotation, like "said", "that", "why", and other reporting or questioning language.

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

 kasarigan

-0.72

 AppCompatTheme

-0.61

 Meksiku

-0.52

 invokingState

-0.52

Rohy

-0.51

KommentareTeilen

-0.50

libft

-0.48

Datuak

-0.47

 gainera

-0.46

 Normdatei

-0.46

POSITIVE LOGITS

ValueStyle

0.37

 Bucure

0.36

 aparente

0.36

ässä

0.36

 aggiungere

0.36

 ſtill

0.35

 seemingly

0.35

 apparently

0.35

 “…

0.35

 Añade

0.35

Activations Density 0.255%

instances of dialogue or reported speech

words related to speech and quotation, like "said", "that", "why", and other reporting or questioning language.

No Comments

No Known Activations