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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Repeat

-0.71

Fall

-0.65

Document

-0.65

conom

-0.63

Rand

-0.63

Mess

-0.63

Pan

-0.62

cam

-0.62

 Mess

-0.62

mans

-0.62

POSITIVE LOGITS

eret

0.82

ãĥķãĤ©

0.79

ãĥ¼ãĥĨãĤ£

0.73

umbing

0.73

imal

0.72

igo

0.72

ãĤ¨ãĥ«

0.71

ouble

0.70

nosis

0.68

GGGG

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.