INDEX

Explanations

terms related to accuracy and efficiency in various contexts

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

performance metrics and qualities

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

-0.71

<h1>

-0.62

-0.54

Australian

-0.49

-0.48

Local

-0.48

Canadian

-0.45

contained

-0.44

(<

-0.44

“

-0.44

POSITIVE LOGITS

abilidad

0.71

osity

0.71

mability

0.69

eability

0.68

ivity

0.68

 Accuracy

0.67

sivity

0.66

abilité

0.66

esomeness

0.66

genicity

0.66

Activations Density 0.014%

terms related to accuracy and efficiency in various contexts

performance metrics and qualities

No Comments

No Known Activations