INDEX

Explanations

attends to tokens that relate to "Open", "ng", and "BOOST" from their respective following tokens

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.06

1:0.08

2:0.04

3:0.08

4:0.15

5:0.46

6:0.04

7:0.06

Negative Logits

 requestCode

-0.27

PathVariable

-0.25

uk

-0.24

migrationBuilder

-0.24

SwitchCompat

-0.24

DataMember

-0.24

antMatchers

-0.23

ment

-0.23

yl

-0.23

ab

-0.22

POSITIVE LOGITS

 Theſe

0.44

 ExecuteAsync

0.44

 bezeichneter

0.42

 للاسماء

0.42

 itſelf

0.42

 Anſ

0.41

ंदीखरीदारी

0.41

 myſelf

0.40

CloseOperation

0.40

 ―――――

0.40

Activations Density 7.772%

attends to tokens that relate to "Open", "ng", and "BOOST" from their respective following tokens

No Comments

No Known Activations