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

0:0.09

1:0.07

2:0.08

3:0.08

4:0.07

5:0.08

6:0.09

7:0.07

8:0.08

9:0.07

10:0.09

11:0.09

Negative Logits

guards

-2.04

weak

-2.01

erous

-1.96

.」

-1.85

?」

-1.79

 interchange

-1.79

vir

-1.78

layer

-1.77

swing

-1.70

 Guards

-1.70

POSITIVE LOGITS

Released

1.75

DEM

1.69

 satell

1.69

 Luxem

1.66

BRE

1.61

 Belgium

1.53

1.49

EUR

1.47

 Interstellar

1.47

 Cohen

1.46

Activations Density 0.000%

No Known Activations

This feature has no known activations.