INDEX

Explanations

references to testimony and legal proceedings

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

information about specific details

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

expandindo

-0.47

 ModelExpression

-0.44

ⓧ

-0.41

Ingredientes

-0.40

ésult

-0.40

Controllo

-0.39

 szö

-0.39

例文帳に追加

-0.39

Portail

-0.39

ITERATURE

-0.39

POSITIVE LOGITS

󠁴

0.46

AndEndTag

0.45

Referencies

0.44

:✨

0.42

apimachinery

0.42

TagMode

0.42

tangentMode

0.42

↩

0.39

httphttps

0.38

 bezeichneter

0.37

Activations Density 0.052%

references to testimony and legal proceedings

information about specific details

No Comments

No Known Activations