INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ingred

-0.85

querque

-0.84

theless

-0.83

catentry

-0.77

fmt

-0.75

 livest

-0.66

 streng

-0.65

 Wiki

-0.62

Origin

-0.62

Official

-0.61

POSITIVE LOGITS

istine

0.73

cock

0.70

hang

0.64

 Rolls

0.61

chin

0.60

 Tories

0.60

erenn

0.59

weights

0.58

du

0.58

rich

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.