INDEX

Explanations

references to individuals or entities believed to be morally problematic or controversial

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

2:0.20

3:0.05

4:0.15

5:0.04

6:0.04

7:0.14

8:0.05

9:0.06

10:0.08

11:0.11

Negative Logits

 rails

-1.81

rio

-1.61

matter

-1.44

enc

-1.44

ebook

-1.40

ohyd

-1.39

rou

-1.38

 selves

-1.37

inho

-1.34

iour

-1.34

POSITIVE LOGITS

bilt

2.02

��

1.92

龍契士

1.75

bid

1.65

 Azerbai

1.65

ategory

1.64

Zup

1.57

reditation

1.53

actionDate

1.52

��

1.52

Activations Density 0.002%

references to individuals or entities believed to be morally problematic or controversial

No Comments

No Known Activations