INDEX

Explanations

attends to project-related tokens from various related tokens earlier in the sequence

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

0:0.36

1:0.17

2:0.12

3:0.06

4:0.08

5:0.02

6:0.07

7:0.08

Negative Logits

.*")]

-0.35

AnimationsModule

-0.30

IsContent

-0.29

 الحره

-0.28

SequentialGroup

-0.28

localctx

-0.28

polated

-0.28

]--;

-0.27

InputBorder

-0.26

ricing

-0.26

POSITIVE LOGITS

SharedCtor

0.24

 rada

0.19

 colectiva

0.19

Источник

0.18

enderror

0.17

<%@

0.17

jupiter

0.17

 voci

0.17

 Bode

0.17

 nurt

0.16

Activations Density 0.476%

attends to project-related tokens from various related tokens earlier in the sequence

No Comments

No Known Activations