INDEX

Explanations

references to news organizations and detailed news content

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

references to news organizations and reports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-att-kk

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

0:0.05

1:0.06

2:0.03

3:0.04

4:0.03

5:0.05

6:0.08

7:0.15

8:0.04

9:0.15

10:0.18

11:0.09

Negative Logits

 gradient

-1.11

 reversible

-1.04

scale

-0.98

 discard

-0.90

abase

-0.89

 groove

-0.88

 coerc

-0.87

ministic

-0.87

 Yamato

-0.86

example

-0.85

POSITIVE LOGITS

Politics

1.19

News

1.12

 Watergate

1.10

Media

1.04

 Journalism

1.03

 journalism

1.00

PHOTOS

0.97

CNN

0.95

arthed

0.95

 Benghazi

0.94

Activations Density 0.356%

references to news organizations and detailed news content

references to news organizations and reports

No Comments

No Known Activations