INDEX

Explanations

expressions related to conflict and fighting

oai_token-act-pair · gpt-3.5-turbo

references to situations or conditions that are categorized as "full-blown" crises or conflicts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

Ń·

-0.86

Pok

-0.76

 Tycoon

-0.75

 charms

-0.68

Phi

-0.67

Zar

-0.67

 Reloaded

-0.66

 Springs

-0.66

Tik

-0.64

imaru

-0.64

POSITIVE LOGITS

fledged

1.36

bodied

1.23

sized

1.20

blown

1.14

function

1.07

scale

1.05

empt

1.02

hearted

1.02

estate

1.02

circ

1.01

Activations Density 0.029%

expressions related to conflict and fighting

references to situations or conditions that are categorized as "full-blown" crises or conflicts

No Comments

No Known Activations