INDEX

Explanations

references to specific episodes

oai_token-act-pair · gpt-3.5-turbo

references to specific episodes of television shows

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ntil

-0.90

milo

-0.80

hee

-0.80

sheets

-0.76

wegian

-0.73

enei

-0.71

pse

-0.71

gn

-0.70

 unchecked

-0.69

covered

-0.69

POSITIVE LOGITS

 episode

1.27

 episodes

1.06

Episode

0.92

episode

0.91

 Episode

0.89

 finale

0.86

isodes

0.86

ĸļ

0.84

 opener

0.76

ODE

0.75

Activations Density 0.014%

references to specific episodes

references to specific episodes of television shows

No Comments

No Known Activations