INDEX

Explanations

mentions of the word "ve"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

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

ve

-2.50

Ve

-1.57

Ve

-1.35

VE

-1.13

ve

-0.85

VE

-0.85

 veer

-0.72

EndProject

-0.70

ver

-0.67

 Veil

-0.65

POSITIVE LOGITS

 anſ

0.58

 occaf

0.55

 perfons

0.52

 Cæsar

0.52

 noexcept

0.51

ContentLoaded

0.51

 doubtnut

0.50

 autoradio

0.50

 ſche

0.49

yordu

0.49

Activations Density 0.206%

mentions of the word "ve"

No Comments

No Known Activations