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

1:0.07

2:0.09

3:0.08

4:0.09

5:0.07

6:0.07

7:0.08

8:0.08

9:0.07

10:0.07

11:0.09

Negative Logits

inventoryQuantity

-1.80

 chromosomes

-1.53

ournals

-1.52

 compass

-1.50

 attracts

-1.47

ocally

-1.46

 drinkers

-1.44

 Volunte

-1.44

 gastro

-1.43

 Firefly

-1.43

POSITIVE LOGITS

osi

1.77

raq

1.53

Mur

1.52

astical

1.46

 procedural

1.44

Wyr

1.41

Pad

1.40

 anew

1.40

 slashing

1.40

Sut

1.40

Activations Density 0.000%

No Known Activations

This feature has no known activations.