INDEX

Explanations

keywords related to television series, especially season finales

oai_token-act-pair · gpt-3.5-turbo

references to television series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

stadt

-0.77

zie

-0.76

prus

-0.75

nai

-0.74

orgetown

-0.73

ria

-0.73

zan

-0.72

selves

-0.70

ntil

-0.70

bell

-0.70

POSITIVE LOGITS

 finale

1.23

manship

0.94

 regulars

0.93

 premiere

0.89

hip

0.86

hips

0.85

 opener

0.77

'-

0.76

ĸļ

0.73

Activations Density 0.045%

keywords related to television series, especially season finales

references to television series

No Comments

No Known Activations