INDEX

Explanations

phrases related to dangerous incidents, crimes, and police involvement

oai_token-act-pair · gpt-3.5-turbo

instances of the word "the" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ledge

-0.89

BILITIES

-0.81

imil

-0.77

ipedia

-0.75

 Machina

-0.70

witch

-0.70

 certs

-0.67

helm

-0.67

çļ

-0.66

abe

-0.66

POSITIVE LOGITS

 vicinity

1.42

 meantime

1.33

 aftermath

1.17

 midst

1.10

 incident

1.09

 area

1.08

 absence

1.02

 wake

1.02

 altercation

0.97

 intervening

0.95

Activations Density 0.171%

phrases related to dangerous incidents, crimes, and police involvement

instances of the word "the" in various contexts

No Comments

No Known Activations