INDEX

Explanations

terms related to threats to safety and security, especially in the context of food safety and crime

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

topics related to public health and safety issues

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 undermin

-0.59

 disadvant

-0.55

 bragging

-0.52

 destro

-0.52

 territ

-0.51

 glim

-0.50

 Zucker

-0.50

 sucker

-0.50

 chip

-0.49

 defe

-0.48

POSITIVE LOGITS

issues

0.84

relations

0.82

education

0.72

---

0.71

cars

0.68

formation

0.68

fecture

0.66

attacks

0.66

olitics

0.64

quit

0.62

Activations Density 0.023%

terms related to threats to safety and security, especially in the context of food safety and crime

topics related to public health and safety issues

No Comments

No Known Activations