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

1:0.06

2:0.06

3:0.09

4:0.07

5:0.09

6:0.07

7:0.07

8:0.09

9:0.08

10:0.09

11:0.08

Negative Logits

��

-2.33

priority

-1.86

uggest

-1.73

 strengths

-1.73

success

-1.72

effective

-1.72

quality

-1.71

centered

-1.69

depend

-1.65

temp

-1.64

POSITIVE LOGITS

 Imran

1.97

 Nare

1.92

 Barber

1.89

edin

1.86

 Jinn

1.86

 Edwin

1.80

emonium

1.80

 Schumer

1.78

 Collins

1.78

 Elias

1.76

Activations Density 0.000%

No Known Activations

This feature has no known activations.