INDEX

Explanations

words related to preventing or avoiding negative outcomes or situations

oai_token-act-pair · gpt-3.5-turbo

phrases related to prevention of negative outcomes or events

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

soDeliveryDate

-0.85

 largeDownload

-0.82

å¦

-0.77

REL

-0.76

çľ

-0.66

ĪĴ

-0.65

ãģ®éŃĶ

-0.64

æĪ¦

-0.62

 Compass

-0.61

yi

-0.61

POSITIVE LOGITS

 altogether

0.85

 invasive

0.83

 unnecessary

0.83

 unwanted

0.81

 harmful

0.80

 harm

0.80

 congestion

0.76

 lockout

0.75

 buildup

0.74

 pitfalls

0.73

Activations Density 0.303%

words related to preventing or avoiding negative outcomes or situations

phrases related to prevention of negative outcomes or events

No Comments

No Known Activations