INDEX

Explanations

attends to specific tokens from corresponding entities that indicate a relationship between the two

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-att-16k

Configuration

google/gemma-scope-2b-pt-att/layer_10/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.11

1:0.15

2:0.10

3:0.06

4:0.05

5:0.03

6:0.09

7:0.38

Negative Logits

Sucesor

-0.45

verwijspagina

-0.40

 vPvB

-0.39

tonode

-0.36

FFIX

-0.36

 Normdatei

-0.35

chtenstein

-0.35

 NSCoder

-0.34

enken

-0.34

WriteBarrier

-0.34

POSITIVE LOGITS

)";

0.34

})));

0.32

istic

0.31

 Carlson

0.31

Sf

0.30

MinWidth

0.30

οι

0.29

"]();

0.29

 Quicks

0.28

tione

0.28

Activations Density 2.278%

attends to specific tokens from corresponding entities that indicate a relationship between the two

No Comments

No Known Activations