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

0:0.06

1:0.07

2:0.09

3:0.08

4:0.07

5:0.08

6:0.09

7:0.08

8:0.09

9:0.08

10:0.07

11:0.07

Negative Logits

 Models

-1.65

 Casting

-1.64

 obser

-1.62

 Classics

-1.61

 Bret

-1.59

displayText

-1.58

 Tutorial

-1.58

 Generations

-1.54

 scouting

-1.53

 Techniques

-1.51

POSITIVE LOGITS

deen

1.88

ガ

1.46

aintain

1.46

 cushion

1.45

mone

1.45

uably

1.44

 foster

1.41

chain

1.41

car

1.40

archive

1.39

Activations Density 0.000%

No Known Activations

This feature has no known activations.