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

emi

-0.72

Mun

-0.68

 Hebdo

-0.67

 expired

-0.62

 congratulations

-0.61

tained

-0.61

 Moreno

-0.60

went

-0.59

 Wasserman

-0.59

Too

-0.59

POSITIVE LOGITS

ividual

0.75

umblr

0.73

uminati

0.71

 locate

0.70

berra

0.69

 tremend

0.68

isode

0.68

...)

0.66

å§«

0.66

sbm

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.