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

2:0.09

3:0.08

4:0.08

5:0.08

6:0.06

7:0.07

8:0.09

9:0.09

10:0.08

11:0.08

Negative Logits

Medium

-1.84

ciples

-1.70

Intel

-1.69

Micro

-1.68

quad

-1.66

�

-1.62

middle

-1.62

prints

-1.61

Meg

-1.61

Mods

-1.61

POSITIVE LOGITS

 suspended

1.86

oked

1.82

itialized

1.75

 depleted

1.73

umbn

1.71

 downed

1.68

 administered

1.63

 rated

1.56

 condemned

1.55

 administ

1.55

Activations Density 0.000%

No Known Activations

This feature has no known activations.