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

0:0.08

1:0.07

2:0.08

3:0.09

4:0.08

5:0.07

6:0.07

7:0.08

8:0.07

9:0.08

10:0.08

11:0.09

Negative Logits

cellaneous

-1.45

 optionally

-1.45

icut

-1.31

 clauses

-1.30

+/-

-1.24

 typo

-1.23

 cosmetic

-1.23

 ported

-1.22

\'

-1.21

 denotes

-1.21

POSITIVE LOGITS

;"

1.39

 outreach

1.38

acons

1.37

"—

1.31

��

1.27

href

1.26

τ

1.26

NRS

1.26

Merit

1.26

��

1.25

Activations Density 0.000%

No Known Activations

This feature has no known activations.