INDEX

Explanations

elements related to data structures and programming conventions, particularly within a software context

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

items starting with `free`

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

த்த

-0.32

InjectMocks

-0.32

 converted

-0.31

openqa

-0.31

sp

-0.30

 affected

-0.28

NoSuch

-0.28

 conversion

-0.28

und

-0.27

監修

-0.27

POSITIVE LOGITS

NameInMap

0.72

CppMethod

0.72

StructEnd

0.71

 Мексичка

0.66

Aholisi

0.65

رشف

0.65

 transfieras

0.63

 <<<<<<<<<<<<<<

0.62

 tiguan

0.62

ंदीखरीदारी

0.62

Activations Density 0.251%

elements related to data structures and programming conventions, particularly within a software context

items starting with `free`

No Comments

No Known Activations