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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Murray

-0.71

NEC

-0.66

 Murray

-0.65

ELF

-0.64

 Swanson

-0.64

 Processor

-0.63

************

-0.63

ching

-0.63

ched

-0.61

 Examiner

-0.61

POSITIVE LOGITS

hower

1.08

ingen

0.78

udi

0.77

ashtra

0.77

obar

0.68

azard

0.67

acular

0.67

Ì¶

0.66

wagen

0.66

izons

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.