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

lum

-0.71

hex

-0.71

 Chilean

-0.64

encia

-0.64

Interstitial

-0.64

mes

-0.63

 Venezuelan

-0.62

Tonight

-0.62

spl

-0.61

 chili

-0.60

POSITIVE LOGITS

gate

0.71

 breaches

0.71

Raven

0.69

payers

0.69

orthy

0.66

Interested

0.64

Marginal

0.64

uploads

0.64

geries

0.63

ucket

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.