INDEX

Explanations

attends to instances of the token "0" from preceding tokens marked with the function of editing or authorship

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

2:0.59

3:0.09

4:0.04

5:0.02

6:0.02

7:0.07

Negative Logits

IGraphics

-0.42

MemoryWarning

-0.42

 AppDelegate

-0.41

styleUrls

-0.39

Архівовано

-0.38

csname

-0.36

 Schlä

-0.35

 Else

-0.35

Abp

-0.34

UTELY

-0.34

POSITIVE LOGITS

ImageContext

0.41

 steder

0.37

oredCriteria

0.36

SerializedSize

0.35

 يتيمه

0.34

 infância

0.33

 gynhyrchwyd

0.33

lardır

0.33

 душ

0.32

gridx

0.32

Activations Density 1.414%

attends to instances of the token "0" from preceding tokens marked with the function of editing or authorship

No Comments

No Known Activations