INDEX

Explanations

references to specific historical and political events or concepts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

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

Not in Any Lists

No Comments

Head Attr Weights

0:0.03

1:0.03

2:0.11

3:0.03

4:0.41

5:0.04

6:0.02

7:0.03

8:0.05

9:0.11

10:0.06

11:0.02

Negative Logits

 Valve

-1.40

<[

-1.33

nea

-1.33

 Rove

-1.27

 Macron

-1.25

 elector

-1.19

chel

-1.19

nery

-1.18

culus

-1.16

 Drift

-1.15

POSITIVE LOGITS

 EDITION

1.74

 Annotations

1.46

GGGG

1.41

hooting

1.40

paragraph

1.35

姫

1.33

pmwiki

1.31

 Eleven

1.28

LECT

1.27

神

1.27

Activations Density 0.003%

references to specific historical and political events or concepts

No Comments

No Known Activations