INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ilated

-0.70

ocry

-0.66

Lux

-0.66

apse

-0.66

rique

-0.63

uties

-0.63

ipers

-0.63

avi

-0.63

aires

-0.61

wen

-0.61

POSITIVE LOGITS

DragonMagazine

0.85

ãĤ´ãĥ³

0.82

pmwiki

0.73

DCS

0.72

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.69

NAS

0.67

"""

0.67

theless

0.66

ppa

0.64

isSpecialOrderable

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.