INDEX

Explanations

references to monthly reports or documents regarding compensation or remuneration in organizational contexts

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

Head Attr Weights

0:0.17

1:0.02

2:0.28

3:0.08

4:0.02

5:0.13

6:0.03

7:0.04

8:0.09

9:0.03

10:0.03

11:0.02

Negative Logits

DSM

-3.50

OCD

-3.16

 euth

-3.15

 KILL

-3.12

 Creep

-3.05

 Suicide

-3.01

 Shelter

-3.01

 suicidal

-3.00

furt

-2.98

 Population

-2.90

POSITIVE LOGITS

 impe

3.17

 presidents

2.96

 dishon

2.96

 treasury

2.94

 royalty

2.85

 propri

2.77

 generals

2.75

 emperor

2.74

 bribery

2.72

 administrations

2.71

Activations Density 0.000%

references to monthly reports or documents regarding compensation or remuneration in organizational contexts

No Known Activations