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

 stash

-0.75

ilk

-0.73

 Cors

-0.61

 jack

-0.60

pload

-0.58

 Czech

-0.58

 stack

-0.58

ilitarian

-0.58

 Wired

-0.58

 (Â£

-0.58

POSITIVE LOGITS

putable

0.72

Ĥİ

0.69

osate

0.68

aky

0.68

terness

0.65

boards

0.64

 Architect

0.63

erenn

0.62

ĨĴ

0.60

 Signs

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.