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

 Merrill

-0.67

 Merchant

-0.64

 Digest

-0.63

 Monk

-0.62

 Slayer

-0.61

 attendant

-0.60

 Hague

-0.59

 Gutenberg

-0.59

 Premiership

-0.59

 Arcane

-0.58

POSITIVE LOGITS

ividual

0.82

osc

0.79

Characters

0.79

ns

0.77

ibo

0.75

Versions

0.73

riad

0.72

nai

0.72

lat

0.72

itives

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.