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

1:0.09

2:0.08

3:0.08

4:0.07

5:0.09

6:0.09

7:0.08

8:0.06

9:0.07

10:0.08

11:0.08

Negative Logits

 Weekend

-1.52

quished

-1.51

ettings

-1.48

 Cance

-1.44

ignt

-1.43

 Expert

-1.36

 Assist

-1.36

 Assembly

-1.34

 Anthem

-1.34

POSITIVE LOGITS

worms

1.66

aiman

1.65

 bark

1.56

 insecurity

1.49

UTF

1.42

 ripple

1.36

�

1.34

&&

1.33

 contagious

1.33

 uncertain

1.32

Activations Density 0.000%

No Known Activations

This feature has no known activations.