INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 concess

-0.65

 Rahul

-0.63

 appl

-0.62

ipeg

-0.62

 Idle

-0.62

Ping

-0.60

 Kejriwal

-0.60

Uni

-0.59

 Hindi

-0.59

UID

-0.59

POSITIVE LOGITS

agall

0.83

odor

0.76

idy

0.73

opter

0.71

BIL

0.71

alid

0.70

hetics

0.67

 Illusion

0.66

orter

0.65

HAEL

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.