INDEX

Explanations

terms related to barriers or enclosures, particularly in the context of spaces and privacy

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

physical barriers or privacy dividers

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

ContentLoaded

-0.62

 gabbana

-0.45

MemoryWarning

-0.45

 irré

-0.45

 exé

-0.45

arakat

-0.44

 activation

-0.44

 actifs

-0.44

igg

-0.43

Демографія

-0.42

POSITIVE LOGITS

 walls

0.56

 curtain

0.55

 curtains

0.52

 partitions

0.51

 partition

0.47

 Walls

0.47

 wall

0.46

 ComVisible

0.45

 privacy

0.44

ftagPool

0.44

Activations Density 0.009%

terms related to barriers or enclosures, particularly in the context of spaces and privacy

physical barriers or privacy dividers

No Comments

No Known Activations