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

1:0.09

2:0.09

3:0.09

4:0.07

5:0.07

6:0.08

7:0.07

8:0.07

9:0.08

10:0.08

11:0.07

Negative Logits

 ticking

-2.13

 eras

-1.95

 abnorm

-1.78

 incomplete

-1.77

 symbols

-1.77

 outlines

-1.76

Hew

-1.74

cog

-1.71

 distinct

-1.71

fit

-1.69

POSITIVE LOGITS

MRI

2.21

annis

2.12

�醒

2.11

rius

2.10

asin

2.05

conom

2.01

galitarian

2.00

ivan

1.98

1.97

thel

1.96

Activations Density 0.000%

No Known Activations

This feature has no known activations.