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

enne

-0.83

rency

-0.78

paces

-0.72

rosis

-0.72

rity

-0.72

erie

-0.72

acha

-0.70

haus

-0.69

pread

-0.67

 nutrit

-0.66

POSITIVE LOGITS

SourceFile

0.85

ACC

0.74

OWN

0.73

Tube

0.71

PLA

0.70

 Isle

0.67

FTWARE

0.66

 Preferred

0.65

MSN

0.65

Streamer

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.