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

0:0.09

1:0.08

2:0.08

3:0.07

4:0.07

5:0.07

6:0.08

7:0.10

8:0.08

9:0.06

10:0.08

11:0.09

Negative Logits

adjust

-2.57

 scramble

-2.50

 mathemat

-2.46

 reciproc

-2.40

POL

-2.39

 Zurich

-2.38

 Sunny

-2.26

 contag

-2.25

Lew

-2.21

 Irwin

-2.20

POSITIVE LOGITS

album

2.77

aga

2.58

 Manga

2.58

tg

2.51

atto

2.37

emo

2.37

emy

2.36

ector

2.28

hoe

2.28

ega

2.26

Activations Density 0.000%

No Known Activations

This feature has no known activations.