INDEX

Explanations

words related to official positions or roles in organizations

oai_token-act-pair · gpt-3.5-turbo

references to various types of organizational structures, roles, and outcomes

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

ãĥ©ãĥ³

-0.79

ãĥ¼ãĥ«

-0.76

ãĥĥãĥī

-0.68

inventoryQuantity

-0.64

ãĥ¼ãĥ

-0.62

ãĥİ

-0.62

ãĤ¨ãĥ«

-0.61

ãĤ¤

-0.60

Hor

-0.60

Sensor

-0.59

POSITIVE LOGITS

hips

0.82

isode

0.73

eous

0.69

earances

0.69

(<

0.67

 Canter

0.67

.—

0.66

âĢł

0.66

hip

0.65

 convened

0.64

Activations Density 0.591%

words related to official positions or roles in organizations

references to various types of organizational structures, roles, and outcomes

No Comments

No Known Activations