INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 guiActiveUnfocused

-0.77

pod

-0.74

 Clouds

-0.72

 Tree

-0.71

ventures

-0.68

 Seasons

-0.68

 Beautiful

-0.66

 Adventures

-0.65

 Trees

-0.65

 Authors

-0.63

POSITIVE LOGITS

ItemTracker

0.80

ellen

0.76

 steroids

0.74

VD

0.67

 Vanderbilt

0.65

iceps

0.64

 semin

0.63

ancial

0.63

SS

0.62

ilial

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.