INDEX

Explanations

occurrences of the word "sans" and its variations, indicating a focus on that term

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

comic Sans MS, Danshen, Samsungs

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

躇

-0.49

 mères

-0.41

 Pugh

-0.40

omenti

-0.39

PM

-0.38

områ

-0.37

luiting

-0.37

stuk

-0.36

MLLoader

-0.36

område

-0.35

POSITIVE LOGITS

 Sans

2.23

Sans

2.17

 sans

2.17

sans

2.13

 SANS

1.68

 Sanskrit

0.86

 zonder

0.85

san

0.84

 Sams

0.84

SAN

0.82

Activations Density 0.002%

occurrences of the word "sans" and its variations, indicating a focus on that term

comic Sans MS, Danshen, Samsungs

No Comments

No Known Activations