INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Crew

-0.78

Wat

-0.76

 Leone

-0.65

 Haitian

-0.64

 Lunar

-0.64

 Senegal

-0.63

 Trem

-0.63

 Scully

-0.61

 Egyptians

-0.59

 Nile

-0.59

POSITIVE LOGITS

ministic

0.87

insula

0.85

stead

0.82

urbed

0.82

idon

0.81

odox

0.79

cffff

0.78

entious

0.75

umph

0.74

²¾

0.74

Activations Density 0.000%

No Known Activations

This feature has no known activations.