INDEX

Explanations

phrases related to specific entities or groups

oai_token-act-pair · gpt-3.5-turbo

phrases denoting possession or belonging

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

Pg

-0.73

obyl

-0.69

](

-0.65

displayText

-0.65

.''.

-0.64

={

-0.63

 violates

-0.63

foundation

-0.62

 concede

-0.62

mere

-0.61

POSITIVE LOGITS

 choice

1.12

 course

0.97

yes

0.97

 sorts

0.94

 tomorrow

0.94

icial

0.90

 antiquity

0.88

 Choice

0.87

 Tomorrow

0.87

 WWII

0.85

Activations Density 0.162%

phrases related to specific entities or groups

phrases denoting possession or belonging

No Comments

No Known Activations