INDEX

Explanations

attends to tokens indicating a function or operation from tokens that specify a condition or parameter

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

0:0.11

1:0.13

2:0.09

3:0.13

4:0.13

5:0.05

6:0.14

7:0.18

Negative Logits

^(@)

-0.44

ihnachten

-0.39

 ujednoznacz

-0.39

udadera

-0.38

ecute

-0.38

chuhe

-0.37

cifix

-0.37

beiten

-0.37

 препратки

-0.36

adeloupe

-0.35

POSITIVE LOGITS

NameInMap

0.42

GraphicsUnit

0.29

“

0.29

contentLoaded

0.27

AspNetCore

0.27

 myster

0.26

protoimpl

0.26

0.25

 Meksiku

0.25

ineno

0.24

Activations Density 0.087%

attends to tokens indicating a function or operation from tokens that specify a condition or parameter

No Comments

No Known Activations

attends to tokens indicating a function or operation from tokens that specify a condition or parameter

No Comments

No Known Activations