INDEX

Explanations

references to specific durations of time, especially related to hours

oai_token-act-pair · gpt-3.5-turbo

references to time and duration, particularly related to hours worked or spent

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

dinand

-0.81

imaru

-0.72

LECT

-0.68

BuyableInstoreAndOnline

-0.66

mber

-0.64

bard

-0.63

hess

-0.61

eers

-0.61

emort

-0.60

untled

-0.60

POSITIVE LOGITS

glass

1.32

 commute

0.81

Ago

0.78

long

0.77

endment

0.76

 logged

0.72

hour

0.70

 filibuster

0.67

OND

0.67

Hours

0.66

Activations Density 0.053%

references to specific durations of time, especially related to hours

references to time and duration, particularly related to hours worked or spent

No Comments

No Known Activations