INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

tainment

-0.64

SPONSORED

-0.64

much

-0.64

seed

-0.63

collection

-0.62

estine

-0.62

ndra

-0.62

 Olympia

-0.61

met

-0.60

ploy

-0.60

POSITIVE LOGITS

ULAR

0.72

INAL

0.71

çīĪ

0.66

 bite

0.66

IBLE

0.65

inen

0.65

 Forth

0.63

 Kinn

0.62

 Gore

0.61

CFR

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.