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

awed

-0.71

unci

-0.67

uders

-0.65

ylum

-0.64

 confessions

-0.64

usky

-0.63

integ

-0.62

etimes

-0.62

jug

-0.61

rup

-0.61

POSITIVE LOGITS

OTO

0.76

ALD

0.73

 Mech

0.69

)",

0.69

 Dems

0.67

 Pict

0.66

utterstock

0.66

SPONSORED

0.65

 Airl

0.64

vous

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.