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

1:0.08

2:0.08

3:0.09

4:0.08

5:0.08

6:0.08

7:0.07

8:0.08

9:0.09

10:0.08

11:0.08

Negative Logits

idth

-1.75

 eleph

-1.65

 unlawfully

-1.61

 inappropriately

-1.61

 illegally

-1.60

 licens

-1.58

 assignment

-1.52

approved

-1.50

 authorization

-1.50

,'"

-1.49

POSITIVE LOGITS

unity

1.81

nw

1.74

ury

1.59

 Nanto

1.57

 Cheong

1.54

izens

1.52

将

1.51

izen

1.50

merce

1.50

IENCE

1.49

Activations Density 0.000%

No Known Activations

This feature has no known activations.