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

1:0.13

2:0.08

3:0.07

4:0.07

5:0.07

6:0.06

7:0.08

8:0.08

9:0.08

10:0.08

11:0.08

Negative Logits

 retina

-1.81

RT

-1.69

 Huntington

-1.68

 flats

-1.62

 counterfeit

-1.61

 Storm

-1.56

 Gren

-1.54

 Bern

-1.53

 Surf

-1.52

Fib

-1.51

POSITIVE LOGITS

WithNo

2.30

 advoc

2.21

*/(

2.05

 answ

1.99

 obser

1.81

_>

1.76

assuming

1.72

achus

1.72

 therap

1.69

ヘ

1.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.