INDEX

Explanations

instances of the letter 'D'

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.08

1:0.08

2:0.09

3:0.07

4:0.08

5:0.08

6:0.07

7:0.09

8:0.08

9:0.08

10:0.08

11:0.07

Negative Logits

\">

-3.13

Interstitial

-2.99

Buyable

-2.89

 サーティワン

-2.76

)</

-2.62

\\\\

-2.62

ascript

-2.60

toggle

-2.59

Hide

-2.57

*/(

-2.56

POSITIVE LOGITS

erry

2.89

 Bastard

2.85

ile

2.81

Klu

2.70

 Meier

2.60

Dyn

2.59

 Rush

2.58

 Nazi

2.56

Opp

2.55

ker

2.54

Activations Density 0.000%

instances of the letter 'D'

No Known Activations