INDEX

Explanations

phrases related to professions or areas of expertise

oai_token-act-pair · gpt-3.5-turbo

connections between various groups or entities in discussions

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

ixture

-0.70

ONSORED

-0.70

eanor

-0.69

bley

-0.69

ishment

-0.68

butt

-0.68

nesia

-0.68

 THREE

-0.67

 Alone

-0.67

achu

-0.66

POSITIVE LOGITS

 educators

1.45

 thinkers

1.43

 engineers

1.42

 philosophers

1.41

 technicians

1.38

 mathematic

1.37

 businessmen

1.35

 photographers

1.35

 designers

1.35

 scholars

1.34

Activations Density 0.152%

phrases related to professions or areas of expertise

connections between various groups or entities in discussions

No Comments

No Known Activations