INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Footnote

-0.78

objects

-0.74

#$

-0.72

aten

-0.71

Ct

-0.69

adder

-0.69

prototype

-0.68

 masc

-0.66

oops

-0.66

oids

-0.66

POSITIVE LOGITS

quickShipAvailable

0.85

channelAvailability

0.70

 Mahar

0.67

pour

0.66

 Sail

0.66

 Traffic

0.65

 Solitaire

0.65

ahime

0.64

 Peace

0.63

 Sierra

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.