INDEX

Explanations

keywords related to the element "lead"

oai_token-act-pair · gpt-3.5-turbo

mentions of lead as a substance or topic of discussion

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

cale

-0.71

eatures

-0.71

aves

-0.67

ramid

-0.66

ALLY

-0.63

ikes

-0.62

terday

-0.62

 Remem

-0.62

 symmetry

-0.61

urches

-0.60

POSITIVE LOGITS

better

0.99

ership

0.92

gers

0.85

boards

0.83

 poisoning

0.73

lead

0.72

wig

0.72

hole

0.69

/**

0.68

 negotiator

0.66

Activations Density 0.032%

keywords related to the element "lead"

mentions of lead as a substance or topic of discussion

No Comments

No Known Activations