INDEX

Explanations

attends to the token "much" from the paired "much" tokens in the context of magnitude or intensity

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

0:0.06

1:0.04

2:0.01

3:0.10

4:0.31

5:0.03

6:0.06

7:0.07

8:0.03

9:0.01

10:0.02

11:0.03

12:0.03

13:0.08

14:0.02

15:0.02

Negative Logits

 desconocido

-0.52

 reducido

-0.49

 libremente

-0.48

 cortos

-0.47

 these

-0.47

 alguno

-0.46

 rather

-0.46

 such

-0.46

 ninguno

-0.45

 amarillas

-0.45

POSITIVE LOGITS

 sutures

0.51

 Textiles

0.51

 Shader

0.51

clusal

0.50

 Apoptosis

0.50

 Mammals

0.49

 Classifier

0.48

Payroll

0.48

 <>",

0.48

 Payroll

0.48

Activations Density 0.030%

attends to the token "much" from the paired "much" tokens in the context of magnitude or intensity

No Comments

No Known Activations