INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

lez

-0.87

irtual

-0.82

ixel

-0.82

anooga

-0.75

glas

-0.71

offer

-0.70

 municip

-0.69

ãĤ¤ãĥĪ

-0.69

inav

-0.68

vez

-0.67

POSITIVE LOGITS

 Admir

0.73

Eck

0.69

 Monarch

0.69

 Nost

0.69

Posts

0.66

 worms

0.65

 Lodge

0.64

Sah

0.63

 Races

0.62

 course

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.