INDEX

Explanations

references to waterproof and durable products

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.25

1:0.09

2:0.05

3:0.05

4:0.07

5:0.02

6:0.06

7:0.13

8:0.03

9:0.06

10:0.05

11:0.09

Negative Logits

 Vanguard

-2.73

insula

-2.66

 Chavez

-2.65

 Canaver

-2.62

 Neurolog

-2.58

vale

-2.58

 Haas

-2.58

MRI

-2.57

 transsexual

-2.56

 Neuroscience

-2.54

POSITIVE LOGITS

 waterproof

6.93

 sealing

3.71

 footwear

3.65

 bead

3.59

 submer

3.53

 jackets

3.51

 polymer

3.45

Water

3.33

 durability

3.33

 moisture

3.31

Activations Density 0.002%

references to waterproof and durable products

No Known Activations