INDEX

Explanations

hotels

oai_token-act-pair · gpt-3.5-turbo

mentions of hotels

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

arily

-0.85

ebook

-0.79

lang

-0.75

lr

-0.74

sonian

-0.73

nir

-0.70

ahead

-0.68

tan

-0.67

edly

-0.66

rums

-0.65

POSITIVE LOGITS

 Hotel

1.10

 Plaza

0.83

ZI

0.81

 Room

0.81

 Polo

0.81

 Rooms

0.77

 Club

0.76

wives

0.76

 Towers

0.75

Hot

0.75

Activations Density 0.029%

hotels

mentions of hotels

No Comments

No Known Activations