INDEX

Explanations

coding syntax and programming structure

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

programming code comments

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

DllImport

-0.42

AutoField

-0.42

 iVar

-0.41

subplots

-0.38

yarnpkg

-0.36

 unknowingly

-0.35

 utafitiHapana

-0.35

Ӕ

-0.35

 yearly

-0.34

 définitivement

-0.34

POSITIVE LOGITS

GOTREF

0.54

TintMode

0.52

verwijspagina

0.49

 Ceinture

0.48

isateur

0.48

WithIOException

0.47

yntaxException

0.46



0.46

tanleria

0.46

 estekak

0.45

Activations Density 0.003%

coding syntax and programming structure

programming code comments

No Comments

No Known Activations