INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

VICE

-0.68

videos

-0.68

 Crew

-0.67

 Logged

-0.67

Developer

-0.66

DS

-0.66

MFT

-0.65

bors

-0.65

 Elon

-0.64

eligible

-0.63

POSITIVE LOGITS

 poisoning

0.80

 satur

0.75

 casc

0.74

 wounding

0.73

luaj

0.73

 trillions

0.72

ultane

0.69

 inhib

0.69

 unchecked

0.67

atical

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.