INDEX

Explanations

occurrences of the word "end" in various contexts

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

closing code blocks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 itſelf

-0.83

$_(

-0.82

 MainAxisSize

-0.79

 snippetHide

-0.78

 queſta

-0.78

Diweddarwch

-0.77

 المعيارى

-0.75

 wikipagina

-0.75

 ſich

-0.75

 Monfieur

-0.74

POSITIVE LOGITS

↵↵

0.99

↵

0.70

↵↵↵

0.69

0.56

↵↵↵↵

0.54

0.50

End

0.49

end

0.48

0.47

0.44

Activations Density 0.007%

occurrences of the word "end" in various contexts

closing code blocks

No Comments

No Known Activations