INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.9.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.07

1:0.08

2:0.08

3:0.08

4:0.09

5:0.06

6:0.07

7:0.07

8:0.07

9:0.07

10:0.11

11:0.10

Negative Logits

-3.19

»

-3.00

•

-2.78

Reviewed

-2.69

Discuss

-2.62

 Rhodes

-2.57

ダ

-2.47

-2.46

rawler

-2.46

-2.45

POSITIVE LOGITS

 mist

2.40

 mistaken

2.39

wic

2.34

 meal

2.30

 whim

2.26

 secondly

2.25

 chop

2.23

 Ninth

2.20

 Nasa

2.16

oku

2.15

Activations Density 0.000%

No Known Activations

This feature has no known activations.