INDEX

Explanations

various forms of the word "Ho" or related phonetic patterns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

HtmlAttribute

-0.67

 الدولى

-0.63

TintMode

-0.61

(/[

-0.55

ViewFeatures

-0.55

matmul

-0.53

 primary

-0.52

 Kuchen

-0.52

URIComponent

-0.52

GEBURTSDATUM

-0.52

POSITIVE LOGITS

ho

0.84

 Hoss

0.81

 Hoek

0.78

Ho

0.77

 Hollis

0.73

HO

0.73

 Hock

0.73

Ho

0.71

 hopper

0.67

Hos

0.67

Activations Density 0.080%

various forms of the word "Ho" or related phonetic patterns

No Comments

No Known Activations