INDEX

Explanations

elements related to programming and graph visualization

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

rendering visuals

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

Personendaten

-0.60

 resourceCulture

-0.52

 cref

-0.50

awaiter

-0.48

findpost

-0.47

 ModelExpression

-0.47

complexType

-0.45

CompleteListener

-0.44

expandindo

-0.43

 Normdatei

-0.42

POSITIVE LOGITS

 visualizations

0.54

 visuel

0.53

 visuals

0.52

 VISUAL

0.52

 visualization

0.52

 visual

0.50

visual

0.50

Visual

0.49

 visualización

0.48

 étoit

0.47

Activations Density 0.022%

elements related to programming and graph visualization

rendering visuals

No Comments

No Known Activations