INDEX

Explanations

attends to the token "pull" from associated directional tokens, indicating various actions of pulling

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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.05

1:0.03

2:0.03

3:0.02

4:0.59

5:0.17

6:0.02

7:0.05

Negative Logits

Efq

-0.59

RegressionTest

-0.57

 myſelf

-0.54

ſelves

-0.52

 HasFactory

-0.52

 nahilalakip

-0.52

Hochspringen

-0.51

 متعلقه

-0.50

 ―――――

-0.50

ViewFeatures

-0.49

POSITIVE LOGITS

0.31

 членов

0.25

рных

0.25

 môn

0.25

na

0.24

 없습니다

0.24

рного

0.24

декса

0.24

acağı

0.24

Activations Density 0.026%

attends to the token "pull" from associated directional tokens, indicating various actions of pulling

No Comments

No Known Activations