INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ega

-0.75

etheless

-0.72

undo

-0.72

onet

-0.71

uador

-0.67

bom

-0.66

wana

-0.64

uart

-0.61

 Zurich

-0.61

 Constable

-0.60

POSITIVE LOGITS

 expects

0.67

ulates

0.67

hered

0.66

SELECT

0.63

acy

0.63

Apply

0.63

ared

0.62

 Ridley

0.62

aring

0.61

ana

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.