INDEX

Explanations

time-related phrases or events

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the passage of time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

rouse

-0.61

robat

-0.60

 Brach

-0.59

lication

-0.57

 Intelligent

-0.56

PR

-0.55

 Third

-0.55

la

-0.54

Publisher

-0.54

 pose

-0.54

POSITIVE LOGITS

 awhile

1.29

 decades

0.91

 raining

0.91

 years

0.87

 centuries

0.84

 months

0.84

 downhill

0.81

 millennia

0.80

 YEAR

0.79

cffffcc

0.75

Activations Density 0.080%

time-related phrases or events

phrases indicating the passage of time

No Comments

No Known Activations