INDEX

Explanations

phrases indicating deception or fraudulent behavior

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.03

1:0.03

2:0.09

3:0.06

4:0.10

5:0.03

6:0.04

7:0.34

8:0.05

9:0.04

10:0.08

11:0.07

Negative Logits

ascript

-1.69

swer

-1.65

 guid

-1.58

elf

-1.56

fax

-1.53

ysis

-1.51

lene

-1.50

jad

-1.49

deen

-1.48

answered

-1.46

POSITIVE LOGITS

isSpecialOrderable

1.98

 Asia

1.55

 precedent

1.49

 Targ

1.46

 internationally

1.44

 Asians

1.43

 Mens

1.43

 Clouds

1.42

Asia

1.40

 nationally

1.39

Activations Density 0.000%

phrases indicating deception or fraudulent behavior

No Comments

No Known Activations