INDEX

Explanations

attends to exported tokens from tokens that are indicative of definitions or module declarations

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

0:0.07

1:0.09

2:0.11

3:0.10

4:0.08

5:0.04

6:0.17

7:0.30

Negative Logits

NameInMap

-0.47

########.

-0.39

AsUp

-0.38

 linkovi

-0.38

windowFixed

-0.36

省市镇

-0.36

Personensuche

-0.35

Šaltiniai

-0.35

Referencie

-0.35

ANTLR

-0.32

POSITIVE LOGITS

 Stande

0.26

cektir

0.25

ptions

0.25

izable

0.25

müş

0.24

leyebilirsiniz

0.24

eabouts

0.24

 виро

0.23

 роботи

0.23

etheless

0.23

Activations Density 0.052%

attends to exported tokens from tokens that are indicative of definitions or module declarations

No Comments

No Known Activations