INDEX

Explanations

the word "the" following by a noun

oai_token-act-pair · gpt-3.5-turbo

references to "the" along with various current events or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

vity

-0.81

auri

-0.80

mur

-0.78

noon

-0.77

inel

-0.72

ldom

-0.70

staking

-0.70

geoning

-0.69

ouble

-0.67

quire

-0.67

POSITIVE LOGITS

 heck

1.38

 hell

1.22

 fuss

1.19

 fuck

0.93

 future

0.82

 HELL

0.75

 next

0.74

 ramifications

0.70

bies

0.70

 difference

0.70

Activations Density 0.095%

the word "the" following by a noun

references to "the" along with various current events or situations

No Comments

No Known Activations