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

jri

-0.79

credit

-0.76

uly

-0.75

rugged

-0.73

ulia

-0.73

 robbed

-0.67

mia

-0.67

apse

-0.66

ove

-0.66

 pains

-0.66

POSITIVE LOGITS

 GOODMAN

0.80

imeter

0.74

EStream

0.73

imeters

0.68

ãĥ¯

0.68

BILITIES

0.67

BAT

0.66

 sqor

0.66

MENTS

0.66

ADS

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.