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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

elines

-0.73

oxy

-0.62

ook

-0.62

 outl

-0.61

 gospel

-0.61

¥µ

-0.60

adies

-0.59

wegian

-0.59

otics

-0.59

english

-0.57

POSITIVE LOGITS

_-

0.77

delay

0.71

ãĥ»

0.71

 confir

0.67

[*]

0.65

vm

0.64

IUM

0.63

.............

0.63

âĶĢ

0.63

CLASSIFIED

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.