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

 Pinball

-0.76

Ips

-0.74

 Marilyn

-0.69

 Canaver

-0.69

 McMaster

-0.69

 Milton

-0.68

Lex

-0.67

 Geral

-0.66

 legion

-0.65

 Gould

-0.65

POSITIVE LOGITS

ITNESS

0.88

folios

0.85

ratulations

0.81

folio

0.79

fman

0.79

dain

0.78

akeru

0.77

oldown

0.76

inement

0.76

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

0.74

Activations Density 0.000%

No Known Activations

This feature has no known activations.