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

Head Attr Weights

0:0.08

1:0.11

2:0.08

3:0.08

4:0.08

5:0.09

6:0.08

7:0.07

8:0.07

9:0.06

10:0.08

11:0.07

Negative Logits

pter

-1.76

 Ples

-1.60

ctors

-1.56

abase

-1.54

 Resp

-1.54

 Beet

-1.52

 Cutter

-1.49

Aph

-1.47

hoe

-1.44

機

-1.43

POSITIVE LOGITS

 ripple

1.70

 graffiti

1.55

Finish

1.54

mit

1.51

 Romeo

1.50

Hide

1.49

wcs

1.47

Explore

1.47

race

1.45

Arab

1.41

Activations Density 0.000%

No Known Activations

This feature has no known activations.