INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

azine

-0.88

eph

-0.80

oke

-0.77

ãĥĥãĤ¯

-0.74

ovi

-0.72

 registrations

-0.72

Registered

-0.70

cycles

-0.69

agos

-0.68

ibo

-0.68

POSITIVE LOGITS

 Ancients

0.78

 contradiction

0.69

false

0.66

 Moss

0.65

 Behind

0.65

Yet

0.63

 contradictory

0.62

 Sharon

0.61

 barley

0.61

Opp

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.