INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

govtrack

-0.82

zai

-0.76

whe

-0.73

avery

-0.70

aqu

-0.69

Posts

-0.69

rontal

-0.69

 Views

-0.67

ĪĴ

-0.67

ham

-0.66

POSITIVE LOGITS

 cooperation

0.64

ttle

0.64

 detection

0.63

 misfortune

0.63

osit

0.62

WARN

0.60

urion

0.59

OSS

0.59

 obstruction

0.58

auri

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.