INDEX

Explanations

phrases related to plots and orchestrated actions

oai_token-act-pair · gpt-3.5-turbo

references to conspiracy, plotting, or orchestration of actions or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

alli

-0.83

framework

-0.77

esa

-0.75

doi

-0.73

english

-0.73

atha

-0.72

 immedi

-0.70

thur

-0.70

fol

-0.69

aqu

-0.69

POSITIVE LOGITS

 atrocities

1.13

 crimes

1.12

 terrorist

1.10

 murders

1.09

 terrorism

1.09

 murder

1.09

 crime

1.08

 kidnapping

1.07

 sabotage

1.07

 wrongdoing

1.03

Activations Density 0.273%

phrases related to plots and orchestrated actions

references to conspiracy, plotting, or orchestration of actions or events

No Comments

No Known Activations