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

acebook

-0.84

ernel

-0.75

Asset

-0.74

rodu

-0.73

Narr

-0.70

phabet

-0.69

ventory

-0.68

 Shed

-0.68

Veh

-0.67

ocobo

-0.66

POSITIVE LOGITS

 Guerrero

0.90

 Rica

0.74

 glasses

0.70

 Uruguay

0.67

 whist

0.66

 inclusive

0.66

 orient

0.65

istan

0.65

 Guam

0.65

 Asgard

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.