INDEX

Explanations

references to ancient texts or religious documents

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

biblical and technical terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

=’

-0.40

 rare

-0.35

 счет

-0.35

外

-0.35

 crest

-0.34

cherie

-0.33

Vé

-0.33

<_>

-0.32

tea

-0.32

fle

-0.32

POSITIVE LOGITS

 Romans

1.95

Romans

1.82

 romans

1.16

 Hebrews

1.07

Hebrews

1.05

 Corinthians

1.02

 Ephesians

0.92

inthians

0.90

atians

0.87

 romanos

0.87

Activations Density 0.002%

references to ancient texts or religious documents

biblical and technical terms

No Comments

No Known Activations