INDEX

Explanations

mentions or discussions of sequels

oai_token-act-pair · gpt-3.5-turbo

mentions of sequels in various contexts

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

¬¼

-0.72

hoff

-0.65

Ĭ±

-0.64

uda

-0.63

·

-0.61

urrency

-0.61

 Athletics

-0.60

Cola

-0.59

romy

-0.58

 traders

-0.58

POSITIVE LOGITS

 sequel

1.03

 sequels

1.02

ngth

0.91

 trilogy

0.86

arium

0.79

 thereto

0.77

ogue

0.76

 installments

0.75

 installment

0.75

culosis

0.75

Activations Density 0.025%

mentions or discussions of sequels

mentions of sequels in various contexts

No Comments

No Known Activations