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

mo

-0.68

docs

-0.66

mere

-0.64

bent

-0.64

 Chrom

-0.63

 Bian

-0.62

 Arbor

-0.61

mop

-0.61

upload

-0.60

mint

-0.60

POSITIVE LOGITS

HAEL

0.70

 Rounds

0.70

oka

0.69

PDATE

0.67

BLE

0.65

cules

0.64

 calf

0.64

rored

0.63

TED

0.63

actic

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.