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

 bonded

-0.85

 destro

-0.80

cius

-0.74

 Wrong

-0.70

 bonding

-0.70

 impressions

-0.70

ewitness

-0.68

paio

-0.66

aukee

-0.64

 condem

-0.64

POSITIVE LOGITS

SHARE

0.75

é¾į

0.69

AFP

0.69

raft

0.67

Money

0.65

Dat

0.64

BILITIES

0.64

nings

0.63

Access

0.62

 Elections

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.