INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

iol

-0.76

haar

-0.72

fect

-0.70

 certification

-0.68

iral

-0.65

 Certification

-0.64

oshenko

-0.64

andise

-0.64

agine

-0.61

haps

-0.59

POSITIVE LOGITS

river

0.80

UTC

0.76

polit

0.67

 regretted

0.67

smith

0.67

anwhile

0.65

eous

0.65

intosh

0.61

ages

0.61

ãĥ¼ãĥ«

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.