INDEX

Explanations

attends to programming-related tokens from import statements

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

0:0.12

1:0.27

2:0.09

3:0.10

4:0.17

5:0.08

6:0.05

7:0.08

Negative Logits

hmig

-0.43

 $("<

-0.43

 للمعارف

-0.37

 Samo

-0.37

*********/

-0.37

phonie

-0.36

ladimir

-0.36

]='\

-0.36

:✨

-0.36

 ECHO

-0.36

POSITIVE LOGITS

ServletConfig

0.41

 onAnimation

0.40

hilt

0.38

########.

0.36

 zależności

0.35

 Roper

0.35

ьаж

0.34

 '\\;'

0.34

GraphicsUnit

0.33

FontOfSize

0.33

Activations Density 0.018%

attends to programming-related tokens from import statements

No Comments

No Known Activations