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

hement

-0.81

URA

-0.76

ccording

-0.76

 corrid

-0.69

 Shroud

-0.68

wright

-0.67

%"

-0.66

 Buch

-0.65

 arrang

-0.65

iland

-0.65

POSITIVE LOGITS

peak

0.71

 hurricanes

0.68

ox

0.67

ithmetic

0.64

 alum

0.62

gom

0.62

ty

0.60

 grounding

0.60

ttp

0.60

icrobial

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.