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

Reviewer

-0.81

selves

-0.74

HCR

-0.69

hest

-0.69

lings

-0.68

ngth

-0.68

 Gleaming

-0.66

Liter

-0.64

 Representative

-0.63

Footnote

-0.63

POSITIVE LOGITS

oped

0.84

prus

0.76

urat

0.73

aced

0.70

rist

0.67

edIn

0.66

ead

0.63

imar

0.63

aver

0.60

 proxy

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.