INDEX

Explanations

code-related terms that indicate errors or status notifications

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

indicators of errors in network responses

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

%v/v, %w/v, %w/w

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

Referències

-0.61

 تضيفلها

-0.47

 Inner

-0.46

 reembolso

-0.46

Inner

-0.45

IFO

-0.42

RunWith

-0.42

 beso

-0.41

inner

-0.41

POSITIVE LOGITS

 ModelExpression

0.84

})->

0.66

 <<<<<<<<<<<<<<

0.65

"]}

0.64

RTEE

0.59

orice

0.58

__':

0.58

lably

0.57

}{*}{}

0.57

 AssemblyVersion

0.57

Activations Density 0.033%

code-related terms that indicate errors or status notifications

indicators of errors in network responses

%v/v, %w/v, %w/w

No Comments

No Known Activations