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

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.15

2:0.06

3:0.07

4:0.08

5:0.07

6:0.07

7:0.06

8:0.09

9:0.07

10:0.07

11:0.08

Negative Logits

abus

-2.27

ongs

-2.25

ographics

-2.13

onomy

-2.12

baugh

-2.12

ainers

-2.08

gart

-2.01

andra

-1.90

ihad

-1.89

ees

-1.89

POSITIVE LOGITS

 Siberia

1.88

 Leviathan

1.87

 crack

1.66

 somewhere

1.64

 speculated

1.63

 SWAT

1.61

 speculate

1.59

 conjecture

1.59

 Overwatch

1.50

 fodder

1.49

Activations Density 0.000%

No Known Activations

This feature has no known activations.