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

1:0.05

2:0.06

3:0.08

4:0.09

5:0.07

6:0.08

7:0.09

8:0.07

9:0.07

10:0.10

11:0.09

Negative Logits

oiler

-1.93

ain

-1.91

 suffice

-1.75

bre

-1.65

DON

-1.60

rites

-1.57

blo

-1.55

 Berry

-1.52

redients

-1.51

fo

-1.50

POSITIVE LOGITS

yss

1.71

 concurrent

1.58

ultane

1.55

witch

1.50

 interfaces

1.48

icz

1.47

ward

1.44

edom

1.43

 contrad

1.42

uously

1.41

Activations Density 0.000%

No Known Activations

This feature has no known activations.