INDEX

Explanations

markup elements and structured data formats

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

code and markup delimiters

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

GEBURTS

-0.62

 zijne

-0.57

EDEFAULT

-0.56

Jîn

-0.53

featureID

-0.53

 contextLoads

-0.52

isContained

-0.51

 dezelve

-0.50

UpInside

-0.49

linawan

-0.49

POSITIVE LOGITS

BufferException

0.43

rrggbb

0.39

感

0.36

ValueStyle

0.35

矢

0.35

ToolStrip

0.35

ScopeManager

0.33

nameof

0.33

Carreira

0.32

 heel

0.32

Activations Density 0.190%

markup elements and structured data formats

code and markup delimiters

No Comments

No Known Activations