INDEX

Explanations

code blocks and programming constructs related to object manipulation

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

programming keywords and structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

</i>

-0.59

</b>

-0.50

-0.49

</u>

-0.47

</em>

-0.46

↵

-0.45

</code>

-0.44

</h3>

-0.43

...

-0.43

</h4>

-0.41

POSITIVE LOGITS

expandindo

0.68

 disambiguazione

0.63

parsedMessage

0.61

SharedDtor

0.60

setVerticalGroup

0.60



0.59

SourceChecksum

0.57

Portail

0.54

esModule

0.54

 continúas

0.54

Activations Density 0.007%

code blocks and programming constructs related to object manipulation

programming keywords and structures

No Known Activations