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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EEP

-0.68

sonian

-0.67

eas

-0.66

FW

-0.65

RY

-0.65

 Oath

-0.64

 score

-0.62

 Shall

-0.62

Ws

-0.61

vernment

-0.60

POSITIVE LOGITS

 unfocused

0.70

ryu

0.67

 accelerator

0.66

 obscured

0.65

 unarmed

0.64

 Escape

0.61

 Suzanne

0.60

igen

0.60

urion

0.59

 nausea

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.