INDEX

Explanations

code structure and function return statements

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

code blocks and parentheses

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

Not in Any Lists

No Comments

Negative Logits

AddTagHelper

-0.60

agir

-0.47

 BELIEVE

-0.44

endpush

-0.44

</thead>

-0.43

뀐

-0.43

 CURIAM

-0.43

 feltro

-0.41

DoubleQuotes

-0.41

CppMethod

-0.41

POSITIVE LOGITS

InstrumentedTest

0.46

kuuta

0.42

paravant

0.42

AndroidJUnit

0.40

assertRaises

0.39

Jeografia

0.38

[::-

0.38

хьтан

0.37

IsContent

0.37

 utafitiHapana

0.36

Activations Density 0.179%

code structure and function return statements

code blocks and parentheses

No Comments

No Known Activations