INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 dorm

-0.70

SI

-0.67

olean

-0.64

 undergraduate

-0.63

 yield

-0.63

 gravity

-0.63

 furnace

-0.61

ordinate

-0.61

 overthrow

-0.60

 mattress

-0.59

POSITIVE LOGITS

gans

0.75

planet

0.75

Guest

0.73

kamp

0.70

club

0.70

ingen

0.69

drops

0.68

hours

0.66

âĢ¦)

0.65

-)

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.