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

0:0.08

1:0.07

2:0.08

3:0.08

4:0.07

5:0.07

6:0.08

7:0.09

8:0.09

9:0.08

10:0.08

11:0.07

Negative Logits

ubis

-2.97

selage

-2.91

 Sanchez

-2.87

 Lung

-2.68

 Cairo

-2.67

berus

-2.59

Turkey

-2.53

clair

-2.49

anchez

-2.46

atown

-2.45

POSITIVE LOGITS

 Dominion

2.44

 symmetry

2.43

 accumulating

2.34

 protection

2.32

 storage

2.26

iatures

2.26

 endurance

2.23

 doctrine

2.22

 antim

2.22

 durability

2.21

Activations Density 0.000%

No Known Activations

This feature has no known activations.