INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Hedge

-0.69

 reper

-0.62

 timeout

-0.61

 Helic

-0.60

eto

-0.60

 foregoing

-0.59

DG

-0.59

 Helmet

-0.58

 Hawks

-0.58

iT

-0.58

POSITIVE LOGITS

unity

0.83

anu

0.70

abases

0.70

onomy

0.65

Conservative

0.64

ignty

0.64

velength

0.64

ovo

0.63

hene

0.62

ait

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.