INDEX

Explanations

expressions of distrust and suspicion towards institutions and immigration policies

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

Head Attr Weights

0:0.43

1:0.03

2:0.08

3:0.04

4:0.06

5:0.05

6:0.02

7:0.02

8:0.09

9:0.04

10:0.04

11:0.04

Negative Logits

illac

-1.69

↵

-1.67

rb

-1.61

 helicop

-1.61

cpu

-1.59

 operates

-1.58

 Debor

-1.56

 IMAGES

-1.55

 Baal

-1.53

 Cantor

-1.51

POSITIVE LOGITS

 refusal

2.12

idation

1.93

terness

1.91

itism

1.90

 loss

1.87

lessness

1.86

aturation

1.83

manship

1.82

 absence

1.82

clusion

1.75

Activations Density 0.282%

expressions of distrust and suspicion towards institutions and immigration policies

No Known Activations