INDEX

Explanations

punctuation marks and formatting characters indicating sentences or lists

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

Text after periods

np_acts-logits-general · gemini-2.0-flash

lists punctuation, then words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

 variously

-0.52

 särskilt

-0.51

 offenbar

-0.51

fatter

-0.47

 freilich

-0.47

smallskip

-0.47

 nominally

-0.46

 daarmee

-0.44

あるいは

-0.44

eabouts

-0.44

POSITIVE LOGITS

 AssemblyTitle

0.63

 Lastly

0.61

numerusform

0.57

+#+#

0.56

慢慢的

0.55

 typelib

0.54

Lastly

0.54

 lastly

0.53

不断的

0.53

 utafitiHapana

0.52

Activations Density 0.394%

punctuation marks and formatting characters indicating sentences or lists

Text after periods

lists punctuation, then words

No Comments

No Known Activations