INDEX

Explanations

phrases related to the duration of time or the length of existence

oai_token-act-pair · gpt-3.5-turbo

references to durations or time in relation to ongoing activities or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

catentry

-0.90

}}}

-0.77

asio

-0.74

inki

-0.73

rehend

-0.70

insk

-0.65

#$

-0.64

addy

-0.62

urity

-0.60

hak

-0.60

POSITIVE LOGITS

 since

1.28

since

1.08

 continuously

0.84

 intermitt

0.82

 steadily

0.79

 spor

0.77

 ever

0.73

 professionally

0.72

 longest

0.71

 uninterrupted

0.71

Activations Density 0.139%

phrases related to the duration of time or the length of existence

references to durations or time in relation to ongoing activities or situations

No Comments

No Known Activations