INDEX

Explanations

the presence of specific holders associated with titles or positions

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_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

1:0.01

2:0.07

3:0.06

4:0.08

5:0.04

6:0.09

7:0.38

8:0.02

9:0.04

10:0.09

11:0.05

Negative Logits

 grinding

-1.66

orry

-1.59

ussen

-1.58

 sourcing

-1.58

 Dull

-1.57

anooga

-1.55

 ruth

-1.55

 dealing

-1.50

 discipl

-1.49

isans

-1.48

POSITIVE LOGITS

 patents

1.88

 Membership

1.62

 Album

1.60

AES

1.60

 certificates

1.59

SCP

1.55

DL

1.52

 licences

1.52

 ASCII

1.51

 cards

1.51

Activations Density 0.001%

the presence of specific holders associated with titles or positions

No Comments

No Known Activations