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

2:0.12

3:0.12

4:0.08

5:0.06

6:0.09

7:0.05

8:0.07

9:0.08

10:0.07

11:0.07

Negative Logits

 Resist

-1.76

Wars

-1.71

 Shal

-1.56

 UNCLASSIFIED

-1.55

 Tears

-1.48

 Masquerade

-1.46

study

-1.46

Scar

-1.45

Newsletter

-1.45

[&

-1.44

POSITIVE LOGITS

jri

1.69

1.64

hoe

1.63

�

1.56

 incre

1.56

 grou

1.52

ensibly

1.49

 quickest

1.47

 bragging

1.47

escription

1.45

Activations Density 0.000%

No Known Activations

This feature has no known activations.