INDEX

Explanations

code structure elements, such as method definitions and opening braces

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

code syntax and identifiers

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

 ſche

-0.59

 IFTT

-0.53

 <<<<<<<<<<<<<<

-0.51

ValueStyle

-0.51

 Substitution

-0.49

="@+

-0.49

 tslint

-0.48

 jScrollPane

-0.47

 prompt

-0.47

nasio

-0.47

POSITIVE LOGITS

__":

0.47

__':

0.45

Kjelder

0.45

__':

0.44

 aikaa

0.39

 jezelf

0.37

 zincir

0.37

الدراسه

0.36

__":

0.36

IsMutable

0.35

Activations Density 0.001%

code structure elements, such as method definitions and opening braces

code syntax and identifiers

No Comments

No Known Activations