INDEX

Explanations

copyright statements

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

references to copyright information and publication credits

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-att-kk

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_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.3.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

0:0.02

1:0.05

2:0.11

3:0.21

4:0.02

5:0.02

6:0.06

7:0.12

8:0.07

9:0.18

10:0.03

11:0.07

Negative Logits

flame

-1.47

udi

-1.26

disable

-1.21

tz

-1.16

itiveness

-1.16

dule

-1.15

yang

-1.12

ebted

-1.12

emade

-1.12

chat

-1.10

POSITIVE LOGITS

PDATE

1.27

 insider

1.25

 Palest

1.24

 Insider

1.21

 insiders

1.19

STEM

1.18

 proprietary

1.15

Trend

1.10

EPA

1.08

 GOODMAN

1.08

Activations Density 0.007%

copyright statements

references to copyright information and publication credits

No Comments

No Known Activations