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

peak

-0.73

itone

-0.67

atered

-0.64

ĵ

-0.63

iment

-0.63

hip

-0.62

 usher

-0.62

onso

-0.62

 Cabin

-0.62

bitcoin

-0.61

POSITIVE LOGITS

reads

0.84

urat

0.68

 qualifies

0.66

ecided

0.64

 crashes

0.63

sterdam

0.63

 differe

0.62

oshenko

0.61

nces

0.61

rises

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.