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

orers

-0.80

 cele

-0.63

urtle

-0.62

fram

-0.59

went

-0.59

 unconscious

-0.59

erness

-0.59

untarily

-0.58

compan

-0.58

 autumn

-0.57

POSITIVE LOGITS

Ð´

0.76

lycer

0.71

prototype

0.67

 Chaser

0.66

Dup

0.63

ÄŁ

0.62

MODE

0.61

ãĥĵ

0.61

 Simulation

0.60

ItemThumbnailImage

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.