INDEX

Explanations

statistics related to average values across various categories

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

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

Not in Any Lists

No Comments

Head Attr Weights

0:0.06

1:0.03

2:0.10

3:0.32

4:0.03

5:0.08

6:0.03

7:0.04

8:0.03

9:0.03

10:0.17

11:0.02

Negative Logits

 miracles

-2.34

Names

-2.23

 redund

-2.21

 levers

-2.17

 inventions

-2.07

 antagonists

-2.01

 mysteries

-2.00

 landmarks

-2.00

 innovations

-1.99

farious

-1.98

POSITIVE LOGITS

 average

3.11

average

2.80

 median

2.75

 paycheck

2.71

 income

2.66

 Average

2.62

 averaged

2.55

 salary

2.54

 hourly

2.54

Average

2.54

Activations Density 0.052%

statistics related to average values across various categories

No Comments

No Known Activations