INDEX

Explanations

the presence of the term "additional" in various contexts

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

"additional" and associated data

np_acts-logits-general · gemini-2.0-flash

additional files and data

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 NSCoder

-0.74

󠁮

-0.69

Хьажоргаш

-0.66

 Houſe

-0.65

 Anſ

-0.63

 BorderSide

-0.61

 Reſ

-0.56

 ſtate

-0.56

 IBOutlet

-0.55

ltä

-0.55

POSITIVE LOGITS

 additional

0.66

 subsequent

0.64

nappropriate

0.59

 appropriate

0.59

 sufficient

0.56

 frequently

0.55

 Additional

0.55

 extensive

0.55

Additional

0.55

 subsequently

0.54

Activations Density 0.110%

the presence of the term "additional" in various contexts

"additional" and associated data

additional files and data

No Comments

No Known Activations