INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.3.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

1:0.08

2:0.08

3:0.07

4:0.08

5:0.09

6:0.08

7:0.08

8:0.08

9:0.07

10:0.08

11:0.06

Negative Logits

 viral

-1.49

 antiv

-1.37

 SUPER

-1.36

 psychedel

-1.33

 namely

-1.31

 pertaining

-1.29

branded

-1.28

SER

-1.25

 obnoxious

-1.25

 commercials

-1.23

POSITIVE LOGITS

orget

1.76

giene

1.75

eln

1.65

iciary

1.57

=-=-

1.54

zag

1.52

forts

1.52

agna

1.52

redo

1.52

igsaw

1.51

Activations Density 0.000%

No Known Activations

This feature has no known activations.