INDEX

Explanations

attends to the phrase "of points" from various statements that mention specific tokens related to different contexts

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

0:0.74

1:0.01

2:0.01

3:0.02

4:0.12

5:0.03

6:0.01

7:0.04

Negative Logits

of

-0.34

-0.31

-0.27

by

-0.25

set

-0.24

besos

-0.24

in

-0.23

to

-0.23

 full

-0.22

due

-0.22

POSITIVE LOGITS

StoryboardSegue

0.52

RenderAtEndOf

0.51

 ligiloj

0.47

 дописавши

0.46

Hozzáférés

0.43

ConstraintMaker

0.43

 propOrder

0.42

 tartalomajánló

0.41

IntoConstraints

0.41

 AssemblyProduct

0.40

Activations Density 0.016%

attends to the phrase "of points" from various statements that mention specific tokens related to different contexts

No Comments

No Known Activations