INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

1:0.08

2:0.09

3:0.07

4:0.07

5:0.09

6:0.06

7:0.09

8:0.09

9:0.07

10:0.07

11:0.08

Negative Logits

 Clicker

-1.87

 Reboot

-1.78

 greatness

-1.72

ioxide

-1.68

 Arkham

-1.62

;;;;

-1.60

 SetFontSize

-1.60

 approx

-1.59

+/-

-1.58

 anymore

-1.56

POSITIVE LOGITS

trak

1.91

audi

1.80

interstitial

1.78

ANC

1.65

�

1.64

arter

1.63

田

1.57

nam

1.56

uebl

1.56

iliated

1.54

Activations Density 0.000%

No Known Activations

This feature has no known activations.