INDEX

Explanations

attends to mathematical function or notation tokens from their surrounding context tokens

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

1:0.01

2:0.01

3:0.02

4:0.01

5:0.01

6:0.28

7:0.27

8:0.02

9:0.02

10:0.02

11:0.03

12:0.03

13:0.04

14:0.07

15:0.04

Negative Logits

 avoient

-1.06

 desmotivaciones

-1.02

 étoient

-1.02

ientras

-0.99

 miniaturka

-0.97

 pouvoit

-0.97

 abstrato

-0.97

 étoit

-0.96

ambién

-0.94

 enfans

-0.93

POSITIVE LOGITS

0.93

0.85

0.81

0.78

<eos>

0.76

0.70

Activations Density 1.007%

attends to mathematical function or notation tokens from their surrounding context tokens

No Comments

No Known Activations