INDEX

Explanations

attends to tokens related to "hom" from tokens that are contextually significant in various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_0/width_16k/average_l0_104

Prompts (Dashboard)

36,864 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

0:0.05

1:0.01

2:0.03

3:0.01

4:0.66

5:0.17

6:0.01

7:0.03

Negative Logits

 purpoſe

-0.69

 itſelf

-0.66

 Majefty

-0.66

 raiſ

-0.64

Efq

-0.64

ſelf

-0.63

 ſtate

-0.62

 fubject

-0.62

 neceffary

-0.62

 Jefus

-0.62

POSITIVE LOGITS

}{*}{

0.33

<eos>

0.32

гу

0.30

 nhiêu

0.30

Fonti

0.29

最後に

0.29

im

0.28

ChildScrollView

0.27

pr

0.27

Activations Density 2.675%

attends to tokens related to "hom" from tokens that are contextually significant in various topics

No Comments

No Known Activations