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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 vaccines

-0.82

eva

-0.80

Dak

-0.72

 Vacc

-0.72

 vaccinations

-0.71

 Anthem

-0.70

Nev

-0.69

nikov

-0.69

ratom

-0.68

rament

-0.67

POSITIVE LOGITS

asury

0.81

nown

0.69

 exting

0.69

 guiActiveUn

0.68

 conflic

0.67

UX

0.67

cue

0.66

 livest

0.66

 overflow

0.65

 compos

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.