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

1:0.09

2:0.08

3:0.08

4:0.07

5:0.08

6:0.07

7:0.09

8:0.08

9:0.06

10:0.08

11:0.09

Negative Logits

ashtra

-3.08

sylvania

-2.98

lvl

-2.92

tarian

-2.83

akespe

-2.80

enfranch

-2.69

将

-2.66

uthor

-2.64

aida

-2.57

vot

-2.57

POSITIVE LOGITS

stro

2.98

KC

2.78

stretched

2.70

 Fischer

2.64

 Hicks

2.64

ESC

2.57

AJ

2.53

 Simpson

2.44

 Carlos

2.44

 Nick

2.42

Activations Density 0.000%

No Known Activations

This feature has no known activations.