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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 surg

-0.73

 manned

-0.71

 excavation

-0.70

ittee

-0.69

araoh

-0.69

ledged

-0.68

 nomine

-0.67

ithing

-0.67

cca

-0.66

atives

-0.65

POSITIVE LOGITS

Joy

0.68

Season

0.68

CAR

0.66

car

0.65

Wide

0.65

mb

0.64

CHR

0.64

 curfew

0.64

CTV

0.63

GL

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.