INDEX

Explanations

references to prominent individuals or organizations in political contexts

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

expressions of capability and perceived actions

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

assigning or giving

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

Become

-1.03

 Become

-1.01

 Becoming

-0.99

become

-0.97

Becoming

-0.96

 becoming

-0.96

 become

-0.95

 becomes

-0.94

 Becomes

-0.93

becoming

-0.90

POSITIVE LOGITS

put

0.67

 assign

0.65

 introduce

0.60

 introducing

0.58

oa̍t

0.57

把

0.56

 entrust

0.55

 treating

0.55

 allocate

0.54

把

0.54

Activations Density 2.266%

references to prominent individuals or organizations in political contexts

expressions of capability and perceived actions

assigning or giving

No Comments

No Known Activations