INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

wcsstore

-0.76

arily

-0.73

manac

-0.72

 entitle

-0.70

ogly

-0.69

geries

-0.69

aghan

-0.69

adena

-0.69

uscript

-0.68

 eaves

-0.68

POSITIVE LOGITS

depending

0.83

responsible

0.72

Justice

0.67

for

0.66

SELECT

0.65

while

0.65

via

0.64

 SELECT

0.64

ASH

0.64

CEO

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.