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

Factory

-0.78

¿½

-0.75

 Spartans

-0.75

©¶æ¥µ

-0.69

Fal

-0.69

 Prison

-0.67

Merc

-0.64

MIT

-0.64

 Warden

-0.64

 Fighter

-0.63

POSITIVE LOGITS

orthy

0.85

noon

0.74

ually

0.72

upe

0.70

uality

0.69

 Wales

0.67

alach

0.67

heet

0.66

iliate

0.65

alam

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.