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

bread

-0.89

ãĥ¼ãĤ¯

-0.86

é£

-0.69

ãĤ·ãĥ£

-0.67

ãĥĥ

-0.64

ãĤī

-0.63

Notes

-0.63

çīĪ

-0.62

ores

-0.62

Morning

-0.61

POSITIVE LOGITS

bourg

0.89

soDeliveryDate

0.75

roller

0.66

perature

0.66

ensual

0.66

acterial

0.65

hess

0.63

 lapse

0.63

ply

0.63

 decay

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.