INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_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.9.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

2:0.09

3:0.08

4:0.09

5:0.08

6:0.07

7:0.08

8:0.08

9:0.08

10:0.07

11:0.09

Negative Logits

 Saras

-2.71

rimp

-2.69

 Harding

-2.68

Sew

-2.65

 Nasa

-2.64

skirts

-2.63

Sno

-2.63

 Stockholm

-2.55

 Erie

-2.55

 Swed

-2.54

POSITIVE LOGITS

 Parenthood

2.55

graded

2.48

 redeemed

2.46

itbart

2.44

VERTISEMENT

2.43

cript

2.33

 appre

2.33

needed

2.33

 convert

2.23

 conversion

2.23

Activations Density 0.000%

No Known Activations

This feature has no known activations.