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

2:0.08

3:0.09

4:0.08

5:0.08

6:0.08

7:0.07

8:0.07

9:0.08

10:0.10

11:0.07

Negative Logits

 Flavoring

-1.55

 Associates

-1.53

 Commando

-1.53

��

-1.51

 clos

-1.48

 purs

-1.47

rys

-1.47

FB

-1.46

Ki

-1.46

 scrap

-1.46

POSITIVE LOGITS

vid

1.79

matically

1.71

oke

1.67

 electors

1.64

tracks

1.64

asy

1.61

numbered

1.58

enth

1.56

onal

1.56

estamp

1.54

Activations Density 0.000%

No Known Activations

This feature has no known activations.