INDEX

Explanations

mentions of upcoming seasons of TV shows

oai_token-act-pair · gpt-3.5-turbo

references to television show seasons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gencies

-0.79

hooting

-0.72

leans

-0.71

olated

-0.69

ependent

-0.69

imentary

-0.69

imb

-0.69

diagn

-0.68

 hemorrh

-0.68

arted

-0.65

POSITIVE LOGITS

 finale

1.11

 premiere

1.04

DVD

0.86

 Ultron

0.85

 premie

0.83

 Firefly

0.81

Netflix

0.81

 arcs

0.80

ÃįÃį

0.78

 Sparrow

0.78

Activations Density 0.044%

mentions of upcoming seasons of TV shows

references to television show seasons

No Comments

No Known Activations