INDEX

Explanations

time-related events or deadlines

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "due" indicating scheduled future events or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anon

-0.76

=-=-=-=-=-=-=-=-

-0.68

 Tornado

-0.66

 Islanders

-0.63

 Rampage

-0.63

 sheep

-0.61

 Turtles

-0.61

 Sheep

-0.60

 Column

-0.60

eros

-0.59

POSITIVE LOGITS

 diligence

0.99

lling

0.96

giving

0.82

 cancell

0.77

endment

0.75

ratulations

0.75

ioned

0.74

lled

0.72

 releg

0.72

ration

0.70

Activations Density 0.019%

time-related events or deadlines

occurrences of the word "due" indicating scheduled future events or actions

No Comments

No Known Activations