INDEX

Explanations

errors and exceptions related to programming or computational processes

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

Mathematical notation using symbols and Greek letters

np_acts-logits-general · gemini-2.0-flash

response variable classification

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

}());

-0.57

 algemeen

-0.46

MenuInflater

-0.45

stran

-0.44

 Tenggara

-0.43

ور

-0.41

arrière

-0.41

 verder

-0.41

SourceChecksum

-0.41

輩

-0.41

POSITIVE LOGITS

 referenties

0.74

InitVars

0.71

 EconPapers

0.67

 labels

0.65

ylvan

0.63

ValueStyle

0.61

Datuak

0.61

 truth

0.61

0.60

 gold

0.59

Activations Density 0.626%

errors and exceptions related to programming or computational processes

Mathematical notation using symbols and Greek letters

response variable classification

No Comments

No Known Activations