INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Zot

-0.70

aters

-0.70

 Adin

-0.69

 Parties

-0.67

Services

-0.65

 Nuggets

-0.64

Total

-0.63

lation

-0.63

immers

-0.62

 Tears

-0.62

POSITIVE LOGITS

olean

0.82

 strengths

0.67

 extraord

0.65

iating

0.61

 grandmother

0.60

rises

0.59

 siblings

0.59

 looms

0.59

hom

0.59

aucas

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.