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

 Refugees

-0.66

×

-0.63

Frameworks

-0.63

Compan

-0.62

Intel

-0.62

":["

-0.61

 Kenobi

-0.60

 Jerusalem

-0.59

dimension

-0.58

pires

-0.57

POSITIVE LOGITS

Ħ¢

0.80

 cycle

0.76

olation

0.76

olated

0.69

endars

0.68

 rinse

0.68

ipt

0.67

scill

0.66

 backlog

0.66

 outage

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.