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

Show

-0.70

orsi

-0.70

while

-0.66

Cast

-0.65

Plot

-0.65

terday

-0.64

heimer

-0.64

eri

-0.62

LER

-0.62

Motion

-0.61

POSITIVE LOGITS

usra

0.74

iche

0.69

pillar

0.67

bri

0.64

 sidx

0.63

arcer

0.63

ilater

0.62

ramid

0.62

scl

0.62

 trickle

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.