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

enta

-0.67

yon

-0.67

ographies

-0.66

ourse

-0.65

 Maid

-0.65

iP

-0.65

izon

-0.64

ices

-0.64

osite

-0.63

 Olympia

-0.63

POSITIVE LOGITS

Constructed

0.76

anyahu

0.70

rawdownloadcloneembedreportprint

0.65

NET

0.63

geries

0.63

potion

0.63

runs

0.62

ayan

0.62

Mit

0.61

thodox

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.