INDEX

Explanations

references to specific books and authors

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

written works and fiction

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 CreateTagHelper

-0.44

rungsseite

-0.40

AsUp

-0.40

 çift

-0.37

 allá

-0.36

 Comunicação

-0.36

 Paglinawan

-0.35

 partielle

-0.35

 ordinaria

-0.35

:✨

-0.34

POSITIVE LOGITS

 written

0.56

 novelist

0.52

 novels

0.51

 novel

0.50

 createSlice

0.48

 author

0.47

 bestseller

0.47

 published

0.46

 literary

0.45

 fiction

0.45

Activations Density 0.045%

references to specific books and authors

written works and fiction

No Comments

No Known Activations