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

 coils

-0.75

 cartels

-0.71

 crim

-0.71

 contacts

-0.69

 cooper

-0.69

iasm

-0.69

 infiltrated

-0.66

agon

-0.66

 confiscated

-0.66

 guards

-0.66

POSITIVE LOGITS

Desk

0.79

éĸ

0.73

0.70

BO

0.70

etting

0.69

BOX

0.69

Leary

0.68

birth

0.67

latable

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.