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

Head Attr Weights

0:0.07

1:0.06

2:0.10

3:0.08

4:0.10

5:0.07

6:0.07

7:0.09

8:0.08

9:0.06

10:0.08

11:0.08

Negative Logits

 reckoning

-1.60

 weighing

-1.43

 worlds

-1.41

 alright

-1.40

 stores

-1.40

 temptation

-1.38

 suppliers

-1.37

 ounces

-1.36

 audition

-1.35

 reperto

-1.35

POSITIVE LOGITS

param

1.84

antha

1.71

asar

1.59

arb

1.57

quartered

1.52

ctory

1.49

amon

1.47

ーティ

1.47

dat

1.47

anish

1.46

Activations Density 0.000%

No Known Activations

This feature has no known activations.