INDEX

Explanations

mentions of waiting in line

oai_token-act-pair · gpt-3.5-turbo

references to queues and related concepts

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

oral

-0.81

 mathemat

-0.68

 Arkham

-0.65

inav

-0.64

 advoc

-0.63

aryl

-0.63

sburg

-0.62

ERAL

-0.61

agnetic

-0.60

RAD

-0.60

POSITIVE LOGITS

 queue

1.09

Queue

1.02

 queues

0.99

queue

0.97

erness

0.92

naires

0.91

Que

0.89

que

0.83

naire

0.83

iu

0.80

Activations Density 0.013%

mentions of waiting in line

references to queues and related concepts

No Comments

No Known Activations