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

0:0.09

1:0.07

2:0.08

3:0.09

4:0.08

5:0.07

6:0.09

7:0.07

8:0.09

9:0.08

10:0.08

11:0.07

Negative Logits

 Located

-1.57

authorized

-1.52

 Purchase

-1.46

 Used

-1.39

unless

-1.38

 Except

-1.37

 Petition

-1.36

 downloading

-1.35

 Craigslist

-1.32

 Combine

-1.32

POSITIVE LOGITS

ADRA

1.84

bane

1.73

REDACTED

1.70

iaries

1.65

��

1.63

vic

1.56

anches

1.54

pin

1.53

cffffcc

1.53

 graft

1.52

Activations Density 0.000%

No Known Activations

This feature has no known activations.