INDEX

Explanations

mentions of books or novels

oai_token-act-pair · gpt-3.5-turbo

references to novels or narrative works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ftime

-0.76

halla

-0.73

uler

-0.68

adelphia

-0.67

Sv

-0.65

 Treasurer

-0.62

vals

-0.61

itas

-0.61

gie

-0.61

 Michaels

-0.61

POSITIVE LOGITS

 novel

3.79

 Novel

2.57

 novels

2.47

 novelist

1.74

 screenplay

1.41

 memoir

1.36

 fiction

1.33

 poem

1.32

 manuscript

1.32

 thriller

1.30

Activations Density 0.016%

mentions of books or novels

references to novels or narrative works

No Comments

No Known Activations