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

esis

-0.79

Alice

-0.71

INTER

-0.70

AIDS

-0.70

Kansas

-0.69

inventoryQuantity

-0.69

nor

-0.66

 Rockefeller

-0.65

du

-0.65

Ack

-0.64

POSITIVE LOGITS

PLIED

0.75

xual

0.72

 handlers

0.71

 Marketable

0.70

thouse

0.68

 allowances

0.68

anian

0.67

 licences

0.66

mpeg

0.62

 regimes

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.