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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.10

2:0.07

3:0.07

4:0.08

5:0.09

6:0.08

7:0.08

8:0.10

9:0.07

10:0.07

11:0.06

Negative Logits

Naz

-3.07

 Leopard

-2.83

 Hearthstone

-2.81

 Caucasus

-2.65

Naz

-2.63

 Buddhism

-2.63

Ja

-2.62

Ek

-2.58

 blasphemy

-2.56

Au

-2.53

POSITIVE LOGITS

ply

2.53

 safely

2.49

 demanding

2.48

 stranded

2.45

bill

2.44

 negoti

2.39

wage

2.38

sembly

2.37

 Speedway

2.34

 traveled

2.33

Activations Density 0.000%

No Known Activations

This feature has no known activations.