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

0:0.07

1:0.05

2:0.07

3:0.08

4:0.08

5:0.08

6:0.07

7:0.08

8:0.09

9:0.09

10:0.07

11:0.09

Negative Logits

minist

-1.92

past

-1.82

-1.79

zbollah

-1.76

ernels

-1.73

markets

-1.72

ascus

-1.67

�

-1.67

owder

-1.65

ustain

-1.61

POSITIVE LOGITS

 Participant

1.55

 stranger

1.45

ema

1.43

 Portal

1.40

Gender

1.40

{*

1.37

 voic

1.34

Gig

1.33

 Sapphire

1.31

CG

1.30

Activations Density 0.000%

No Known Activations

This feature has no known activations.