INDEX

Explanations

statements conveying philosophical or intellectual reflections

oai_token-act-pair · gpt-3.5-turbo

references to concepts of significance and importance

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

Alert

-0.98

LED

-0.84

girls

-0.84

poke

-0.83

Tickets

-0.82

nox

-0.79

PHOTOS

-0.77

hops

-0.75

kids

-0.75

HB

-0.75

POSITIVE LOGITS

 notion

1.44

 distinction

1.29

 emergence

1.28

 fallacy

1.27

 doctrine

1.25

 philosopher

1.22

 existence

1.20

 philosophical

1.20

 concept

1.20

 totality

1.18

Activations Density 0.383%

statements conveying philosophical or intellectual reflections

references to concepts of significance and importance

No Comments

No Known Activations