INDEX

Explanations

terms related to various forms of screening, such as medical tests and security checks

oai_token-act-pair · gpt-3.5-turbo

references to the term "screening" 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

rontal

-0.79

lly

-0.72

Else

-0.71

odic

-0.69

rian

-0.69

isin

-0.68

hire

-0.67

acia

-0.66

ciating

-0.66

ordinate

-0.65

POSITIVE LOGITS

 screenings

1.48

 screening

1.35

 screened

1.11

 screens

0.92

 Screen

0.86

TVs

0.78

 screen

0.77

reens

0.76

 Scully

0.74

Screen

0.73

Activations Density 0.023%

terms related to various forms of screening, such as medical tests and security checks

references to the term "screening" in various contexts

No Comments

No Known Activations