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

OUND

-0.79

pir

-0.76

––

-0.75

 AUTHOR

-0.72

âĸĳâĸĳ

-0.70

 +---

-0.69

GREEN

-0.67

VW

-0.64

ships

-0.64

IDER

-0.64

POSITIVE LOGITS

akeru

0.73

 Odin

0.72

 gard

0.71

egal

0.71

hower

0.68

itary

0.68

lus

0.64

opa

0.64

las

0.63

geries

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.