INDEX

Explanations

mentions of containers such as jars or cans, especially those with specific contents or purposes

oai_token-act-pair · gpt-3.5-turbo

references to jars and cans

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 militias

-0.70

voice

-0.69

 Fighters

-0.68

 imperson

-0.68

 streng

-0.66

nom

-0.62

 commute

-0.62

 procedural

-0.62

 defe

-0.61

 antagonist

-0.61

POSITIVE LOGITS

 containing

1.16

 bottles

1.03

 jars

0.99

 contents

0.89

 labeled

0.88

 cans

0.88

jar

0.88

poons

0.88

poon

0.88

lid

0.86

Activations Density 0.202%

mentions of containers such as jars or cans, especially those with specific contents or purposes

references to jars and cans

No Comments

No Known Activations