INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.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

0:0.07

1:0.07

2:0.08

3:0.09

4:0.07

5:0.08

6:0.07

7:0.08

8:0.10

9:0.09

10:0.07

11:0.07

Negative Logits

atus

-1.76

lishes

-1.52

 Crusher

-1.45

ioned

-1.43

onite

-1.40

�士

-1.40

llan

-1.38

Minor

-1.34

 Dupl

-1.30

 mole

-1.29

POSITIVE LOGITS

 afterlife

1.63

 accomplishment

1.51

 entrepreneurship

1.49

tarian

1.43

 anonymity

1.43

 ordeal

1.39

 likeness

1.36

 liking

1.34

 coff

1.33

 embattled

1.31

Activations Density 0.000%

No Known Activations

This feature has no known activations.