INDEX

Explanations

attends to token phrases related to the forms and structures of words from more complex syntactic structures

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.16

1:0.16

2:0.11

3:0.11

4:0.12

5:0.07

6:0.09

7:0.14

Negative Logits

 mxArray

-0.29

تقاوى

-0.24

sơ

-0.24

 Laing

-0.24

cilis

-0.23

gaande

-0.23

ŚĆ

-0.23

حوالہ

-0.22

 SIMILAR

-0.22

 externos

-0.22

POSITIVE LOGITS

TintMode

0.35

DockStyle

0.30

abetes

0.30

EndGlobalSection

0.28

Obrázky

0.28

RequestMapping

0.28

;->

0.28

 OMITBAD

0.27

 &___

0.27

ControllerAdvice

0.26

Activations Density 0.130%

attends to token phrases related to the forms and structures of words from more complex syntactic structures

No Comments

No Known Activations