INDEX

Explanations

phrases related to controversial or ethically questionable actions or events

oai_token-act-pair · gpt-3.5-turbo

terms related to practices, circumstances, and actions in various contexts

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

TTC

-0.69

 RCMP

-0.68

Nun

-0.63

 vodka

-0.63

 bartender

-0.60

 suburbs

-0.60

 Jury

-0.60

 Cure

-0.59

 adolesc

-0.59

Tau

-0.58

POSITIVE LOGITS

 existed

0.91

idious

0.83

azo

0.82

 exists

0.80

roman

0.79

ourning

0.77

ocument

0.74

 exist

0.73

lems

0.73

ãģ®éŃĶ

0.73

Activations Density 0.120%

phrases related to controversial or ethically questionable actions or events

terms related to practices, circumstances, and actions in various contexts

No Comments

No Known Activations