INDEX

Explanations

attends to tokens that indicate compliance or fulfillment of criteria from corresponding tokens that refer to requirements or standards

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

0:0.18

1:0.39

2:0.11

3:0.04

4:0.03

5:0.05

6:0.03

7:0.14

Negative Logits

AddTagHelper

-0.46

AndEndTag

-0.45

 unknownFields

-0.42

 Meksiku

-0.42

 noqa

-0.40

 TestBed

-0.38

@"/

-0.35

InvalidProtocol

-0.34

ształ

-0.34

IntoConstraints

-0.34

POSITIVE LOGITS

naio

0.29

".

0.28

’.

0.27

'.

0.27

”.

0.27

 igjen

0.27

рын

0.27

 sớm

0.26

».

0.26

 subsequently

0.26

Activations Density 0.519%

attends to tokens that indicate compliance or fulfillment of criteria from corresponding tokens that refer to requirements or standards

No Comments

No Known Activations