INDEX

Explanations

code-related structures and syntax, especially related to functions and classes

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

Code-related annotations or serialized names

np_acts-logits-general · gemini-2.0-flash

code constructs and delimiters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.16.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

 esprits

-0.52

 geest

-0.48

 benzer

-0.46

 kesempatan

-0.46

searchResults

-0.45

ław

-0.45

chau

-0.43

 proprement

-0.43

MS

-0.43

 hason

-0.43

POSITIVE LOGITS

IsMutable

0.86

GEBURTSDATUM

0.84

BeginInit

0.78

]")]

0.75

 numberWith

0.73

 defaultstate

0.73

ANTLR

0.73

ThroughAttribute

0.72

hdashline

0.71

/**

0.71

Activations Density 0.215%

code-related structures and syntax, especially related to functions and classes

Code-related annotations or serialized names

code constructs and delimiters

No Comments

No Known Activations