INDEX

Explanations

words indicating a lack of support or denial

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_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

0:0.07

1:0.04

2:0.11

3:0.12

4:0.02

5:0.04

6:0.07

7:0.21

8:0.13

9:0.03

10:0.07

11:0.05

Negative Logits

 Swordsman

-1.38

 Guards

-1.34

 srfAttach

-1.30

 Macro

-1.21

 Pear

-1.17

Magikarp

-1.16

 SetFontSize

-1.13

 Regiment

-1.12

ORGE

-1.11

FontSize

-1.11

POSITIVE LOGITS

ansky

1.34

ramer

1.23

arent

1.17

itate

1.15

agame

1.13

inger

1.13

andom

1.11

lish

1.10

activate

1.08

etsk

1.06

Activations Density 0.033%

words indicating a lack of support or denial

No Comments

No Known Activations