INDEX

Explanations

ordinal numbers in relation to specific positions or rankings

oai_token-act-pair · gpt-3.5-turbo

ordinal numbers and their associated ranks or positions

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

maps

-0.71

matter

-0.70

apt

-0.64

misc

-0.64

iety

-0.61

anced

-0.60

 TEXT

-0.59

ean

-0.59

styles

-0.58

sed

-0.58

POSITIVE LOGITS

 casualty

0.90

 beneficiary

0.79

 reigning

0.76

 beneficiaries

0.76

 hurdle

0.71

liest

0.70

born

0.69

 eligible

0.69

 finalists

0.69

 consecutive

0.68

Activations Density 0.092%

ordinal numbers in relation to specific positions or rankings

ordinal numbers and their associated ranks or positions

No Comments

No Known Activations