INDEX

Explanations

attends to token constructions that include both bold and bracketed numerical indicators from portions of markup code or scientific notation

oai_attention-head · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-att/layer_0/width_16k/average_l0_61

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

0:0.03

1:0.02

2:0.00

3:0.01

4:0.02

5:0.01

6:0.03

7:0.03

8:0.07

9:0.56

10:0.03

11:0.03

12:0.01

13:0.01

14:0.01

15:0.04

Negative Logits

<unused43>

-0.65

<unused8>

-0.64

<unused41>

-0.64

<unused51>

-0.64

<unused16>

-0.64

<unused3>

-0.64

<unused74>

-0.64

<unused42>

-0.64

<unused23>

-0.64

<pad>

-0.64

POSITIVE LOGITS

0.18

0.17

↵

0.15

0.14

 that

0.14

0.13

as

0.13

in

0.13

Activations Density 0.122%

attends to token constructions that include both bold and bracketed numerical indicators from portions of markup code or scientific notation

No Comments

No Known Activations