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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.06

2:0.09

3:0.08

4:0.08

5:0.09

6:0.08

7:0.08

8:0.07

9:0.08

10:0.09

11:0.07

Negative Logits

furt

-1.68

 crashes

-1.53

 srfAttach

-1.53

 looms

-1.48

htaking

-1.47

ultan

-1.46

quartered

-1.45

layer

-1.45

escape

-1.44

 tubes

-1.43

POSITIVE LOGITS

itably

1.58

 acknow

1.57

awa

1.54

Ner

1.51

DCS

1.49

GGGG

1.49

ople

1.48

duino

1.47

 learnt

1.46

 confir

1.45

Activations Density 0.000%

No Known Activations

This feature has no known activations.