INDEX

Explanations

terms related to luring or attracting someone or something

oai_token-act-pair · gpt-3.5-turbo

terms related to bait and deception

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

profits

-0.69

league

-0.64

affer

-0.64

ppard

-0.63

profit

-0.62

 academics

-0.62

Saharan

-0.61

 Universities

-0.60

Rated

-0.60

 Countries

-0.59

POSITIVE LOGITS

 bait

1.50

fish

1.03

ba

0.91

 prey

0.89

 lure

0.86

mong

0.85

glers

0.83

Feed

0.80

plin

0.79

 traps

0.78

Activations Density 0.008%

terms related to luring or attracting someone or something

terms related to bait and deception

No Comments

No Known Activations