INDEX

Explanations

occurrences of the word "Air."

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

air followed by specific words

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

 kesan

-0.40

 gerekti

-0.40

 mümkün

-0.40

싶

-0.40

 Lernen

-0.39

 penampilan

-0.39

 ويكيپيديا

-0.39

bleven

-0.38

 gesetzt

-0.37

期刊论文

-0.37

POSITIVE LOGITS

Air

1.89

Air

1.57

AIR

1.26

AIR

0.97

 Airs

0.82

 Aire

0.79

 AirPods

0.77

 Airtel

0.75

Aire

0.75

 Airborne

0.70

Activations Density 0.003%

occurrences of the word "Air."

air followed by specific words

No Comments

No Known Activations