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

Head Attr Weights

0:0.06

1:0.08

2:0.10

3:0.09

4:0.09

5:0.08

6:0.10

7:0.07

8:0.07

9:0.07

10:0.07

11:0.07

Negative Logits

Gw

-2.60

 Charlotte

-2.50

 Neal

-2.47

 Water

-2.35

CLR

-2.35

iami

-2.33

iverpool

-2.31

 Metropolitan

-2.30

 Plaza

-2.28

 Carmen

-2.26

POSITIVE LOGITS

uten

2.77

repre

2.73

ukemia

2.63

chrom

2.55

 diplom

2.50

tek

2.50

 privile

2.47

 tremend

2.36

 Decay

2.32

 stubborn

2.31

Activations Density 0.000%

No Known Activations

This feature has no known activations.