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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.06

2:0.08

3:0.07

4:0.09

5:0.09

6:0.07

7:0.07

8:0.08

9:0.09

10:0.09

11:0.07

Negative Logits

 ingred

-1.60

 describ

-1.58

hrs

-1.52

 guarant

-1.50

 confir

-1.45

 assets

-1.42

USE

-1.41

duty

-1.41

 condem

-1.38

 allowances

-1.37

POSITIVE LOGITS

 Paran

1.76

quin

1.58

nect

1.55

geist

1.51

phia

1.49

rophe

1.45

lete

1.42

roup

1.42

acet

1.41

lude

1.41

Activations Density 0.000%

No Known Activations

This feature has no known activations.