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

 repetition

-0.71

eden

-0.69

Syd

-0.67

imens

-0.66

İĭ

-0.66

 sample

-0.64

 contemporary

-0.64

ivity

-0.63

 entitle

-0.62

 comprehension

-0.62

POSITIVE LOGITS

sic

0.82

Mesh

0.76

...]

0.75

ONSORED

0.74

REDACTED

0.72

UFF

0.72

?]

0.69

advertising

0.68

FN

0.68

 Maid

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.