INDEX

Explanations

proper nouns and names in the text

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

Names followed by punctuation

np_acts-logits-general · gemini-2.0-flash

names at end of messages

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

SourceChecksum

-0.79

 kasarigan

-0.75

сылкі

-0.74

tvguidetime

-0.71

 كومونز

-0.71

featureID

-0.71

 propOrder

-0.70

nloa

-0.68

ftagPool

-0.66

يكب

-0.66

POSITIVE LOGITS

0.47

↵

0.36

↵↵

0.33

.”

0.32

<!

0.32

..

0.32

↵↵↵

0.32

<eos>

0.31

".

0.30

 Scha

0.29

Activations Density 0.158%

proper nouns and names in the text

Names followed by punctuation

names at end of messages

No Comments

No Known Activations