INDEX

Explanations

No Explanations Found

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

0:0.08

1:0.06

2:0.09

3:0.08

4:0.09

5:0.08

6:0.07

7:0.07

8:0.07

9:0.08

10:0.08

11:0.08

Negative Logits

anus

-1.76

croft

-1.60

 Appearance

-1.58

 Tsukuyomi

-1.58

avia

-1.57

spr

-1.55

 Emblem

-1.52

の�

-1.50

rust

-1.49

OVA

-1.49

POSITIVE LOGITS

versive

1.75

 gent

1.64

 withdrawals

1.62

andise

1.59

 homosexual

1.54

 sermon

1.54

 transact

1.52

 preach

1.52

 sexual

1.52

mercial

1.50

Activations Density 0.000%

No Known Activations

This feature has no known activations.