INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

1:0.08

2:0.07

3:0.09

4:0.09

5:0.08

6:0.07

7:0.07

8:0.07

9:0.07

10:0.09

11:0.09

Negative Logits

wcsstore

-1.77

 rall

-1.75

 ARTICLE

-1.71

edIn

-1.59

 eleph

-1.56

 rapt

-1.56

ilitary

-1.54

 veter

-1.53

 enthusi

-1.52

ahs

-1.52

POSITIVE LOGITS

"""

2.08

agall

1.72

later

1.54

 lying

1.44

lie

1.42

 Banks

1.42

(-

1.40

 thereafter

1.39

 parallel

1.39

*)

1.37

Activations Density 0.000%

No Known Activations

This feature has no known activations.