INDEX

Explanations

attends to tokens marked "official" from various related institutions and entities

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.attn.hook_z

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

Head Attr Weights

0:0.22

1:0.03

2:0.04

3:0.01

4:0.45

5:0.15

6:0.01

7:0.05

Negative Logits

 يتيمه

-0.90

 समीक्षक

-0.80

AndEndTag

-0.78

 myſelf

-0.76

 itſelf

-0.74

 Jefus

-0.71

)";

-0.71

fillType

-0.70

ValueStyle

-0.70

GEBURTSDATUM

-0.70

POSITIVE LOGITS

↵↵

0.52

↵

0.51

0.47

0.46

and

0.44

and

0.44

0.43

<eos>

0.42

—

0.42

in

0.40

Activations Density 0.021%

attends to tokens marked "official" from various related institutions and entities

No Comments

No Known Activations