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

ufact

-0.94

 boycott

-0.81

hesda

-0.75

 boycot

-0.74

 divest

-0.73

osponsors

-0.70

alions

-0.69

orah

-0.69

issions

-0.66

arers

-0.66

POSITIVE LOGITS

Grey

0.71

Tex

0.71

INTON

0.69

Xan

0.68

Gall

0.67

 Giles

0.64

kids

0.62

ouston

0.62

 Dane

0.61

 Avalon

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.