INDEX

Explanations

attends to header tags from corresponding tokens

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

0:0.01

1:0.02

2:0.01

3:0.01

4:0.84

5:0.02

6:0.01

7:0.02

Negative Logits

↵

-0.70

↵↵↵

-0.50

↵↵

-0.47

<strong>

-0.46

-0.44

<h3>

-0.43

</em>

-0.38

...

-0.38

-0.35

-0.34

POSITIVE LOGITS

 itſelf

0.62

 Мексичка

0.57

^(@)

0.57

 AssemblyCulture

0.56

 Roskov

0.54

 Jefus

0.53

 "..\..\..\

0.53

 myſelf

0.52

$_"

0.51

Efq

0.50

Activations Density 0.009%

attends to header tags from corresponding tokens

No Comments

No Known Activations