INDEX

Explanations

concepts and terms associated with political ideologies and discussions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

0:0.03

1:0.01

2:0.04

3:0.05

4:0.01

5:0.04

6:0.01

7:0.04

8:0.02

9:0.01

10:0.65

11:0.02

Negative Logits

 hindsight

-2.38

inventoryQuantity

-2.23

itary

-2.21

needed

-1.98

 impunity

-1.95

warning

-1.91

cember

-1.85

wolves

-1.80

 redacted

-1.78

 doubted

-1.75

POSITIVE LOGITS

 with

4.59

with

4.03

 withd

3.73

 WITH

3.58

 With

3.56

With

3.18

 together

2.19

 dots

2.13

 between

2.10

 alongside

2.06

Activations Density 0.949%

concepts and terms associated with political ideologies and discussions

No Comments

No Known Activations