INDEX

Explanations

dates related to historical events or incidents

oai_token-act-pair · gpt-3.5-turbo

mentions of specific years in a historical context

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

uden

-0.82

abilia

-0.75

phy

-0.74

vier

-0.72

umers

-0.72

wat

-0.72

esses

-0.71

aden

-0.70

alia

-0.70

dain

-0.70

POSITIVE LOGITS

 edition

1.00

 version

0.94

 incarnation

0.94

-'

0.88

 iteration

0.87

 arrival

0.83

 census

0.80

 installment

0.80

 Annual

0.79

 presidential

0.79

Activations Density 0.079%

dates related to historical events or incidents

mentions of specific years in a historical context

No Comments

No Known Activations