INDEX

Explanations

references to someone or something being a "longtime" part of a specific context or community

oai_token-act-pair · gpt-3.5-turbo

the term "longtime" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

abad

-1.00

daq

-0.94

ramid

-0.89

ageddon

-0.88

urgy

-0.87

ammy

-0.87

ibaba

-0.84

apore

-0.82

overe

-0.81

nosis

-0.79

POSITIVE LOGITS

 resident

1.02

 longtime

0.98

 collaborator

0.93

 friend

0.80

 associate

0.76

 ally

0.74

 favorite

0.72

 loyal

0.71

 collaborators

0.70

 member

0.70

Activations Density 0.015%

references to someone or something being a "longtime" part of a specific context or community

the term "longtime" in various contexts

No Comments

No Known Activations