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

2:0.09

3:0.08

4:0.09

5:0.08

6:0.08

7:0.08

8:0.07

9:0.07

10:0.08

11:0.08

Negative Logits

 cheers

-1.69

 Diesel

-1.54

lins

-1.50

 forgot

-1.48

 paperback

-1.48

 roared

-1.45

 gladly

-1.45

bas

-1.43

 wink

-1.42

 moss

-1.41

POSITIVE LOGITS

ournals

2.02

��

1.89

ocumented

1.87

mbuds

1.85

tymology

1.84

hetical

1.72

uments

1.71

ktop

1.70

ilater

1.70

agara

1.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.