INDEX

Explanations

mentions of specific entities or organizations, particularly in the context of economic or political topics

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.04

1:0.02

2:0.17

3:0.05

4:0.13

5:0.10

6:0.04

7:0.06

8:0.08

9:0.06

10:0.12

11:0.07

Negative Logits

 disadvant

-1.39

 inconsist

-1.03

 slippery

-1.01

 psychiat

-0.96

 lawy

-0.96

 juggling

-0.95

 outing

-0.94

romeda

-0.94

ikuman

-0.94

 freeway

-0.94

POSITIVE LOGITS

 guid

1.18

heed

1.06

lycer

1.04

Car

1.02

️

1.02

ortality

1.01

 Allow

1.01

culosis

0.98

Cal

0.97

ascus

0.96

Activations Density 0.294%

mentions of specific entities or organizations, particularly in the context of economic or political topics

No Known Activations