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

1:0.10

2:0.08

3:0.08

4:0.07

5:0.08

6:0.07

7:0.06

8:0.09

9:0.08

10:0.08

11:0.08

Negative Logits

ilus

-1.87

iatrics

-1.65

iband

-1.65

 Redux

-1.63

bah

-1.62

country

-1.60

ctors

-1.60

amina

-1.59

ashtra

-1.59

 Aires

-1.58

POSITIVE LOGITS

VERTISEMENT

1.64

iP

1.59

ittle

1.47

 quar

1.44

 etched

1.44

 swast

1.44

 Pixel

1.41

 alloy

1.38

 Chic

1.37

 Zionism

1.36

Activations Density 0.000%

No Known Activations

This feature has no known activations.