INDEX

Explanations

prominent names or terms associated with historical and cultural references

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

names followed by descriptors

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

fortawesome

-0.28

…

-0.24

起了

-0.23

 marked

-0.23

 aray

-0.23

 Academy

-0.23

 limited

-0.23

 Peach

-0.22

 important

-0.22

 Limited

-0.22

POSITIVE LOGITS

 للاسماء

0.73

GEBURTSDATUM

0.70

 Infórmanos

0.69

 nahilalakip

0.63

oneofs

0.63

 Савезне

0.63

Géographie

0.61

 <<<<<<<<<<<<<<

0.60

fieurs

0.60



0.60

Activations Density 0.047%

prominent names or terms associated with historical and cultural references

names followed by descriptors

No Comments

No Known Activations