INDEX

Explanations

specific institutions and events related to education, sports, and social interactions

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

Head Attr Weights

0:0.01

1:0.02

2:0.09

3:0.06

4:0.43

5:0.01

6:0.03

7:0.16

8:0.02

9:0.03

10:0.05

11:0.03

Negative Logits

expr

-1.72

Chart

-1.70

ipedia

-1.70

ournals

-1.67

adobe

-1.64

chart

-1.56

items

-1.55

Reviewed

-1.54

ilated

-1.52

anguage

-1.52

POSITIVE LOGITS

foe

1.70

 foes

1.68

 incumbent

1.64

 vengeance

1.63

 menace

1.62

 intruder

1.61

 lowly

1.60

 hordes

1.60

 menacing

1.57

 adversity

1.54

Activations Density 0.015%

specific institutions and events related to education, sports, and social interactions

No Known Activations