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

 brill

-0.87

neau

-0.74

 chic

-0.66

 phen

-0.64

 beware

-0.64

 shroud

-0.62

 veil

-0.62

toe

-0.61

å§«

-0.61

venge

-0.60

POSITIVE LOGITS

UNCH

0.75

SHIP

0.73

fuck

0.69

INST

0.68

abus

0.67

Brew

0.66

Ground

0.66

ihu

0.65

jar

0.64

ships

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.