INDEX

Explanations

phrases related to events or conferences and organizational names

oai_token-act-pair · gpt-3.5-turbo

occurrences of geographical locations and events

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

TMZ

-0.90

anuts

-0.90

 goddamn

-0.90

 phony

-0.89

 Bundy

-0.83

 tossed

-0.81

mom

-0.80

 gotten

-0.80

 Sears

-0.80

 paycheck

-0.80

POSITIVE LOGITS

 organis

1.41

 Programme

1.34

 Lect

1.25

 organising

1.25

 programme

1.21

 organise

1.18

 organised

1.17

 organisers

1.15

 Participants

1.15

foc

1.13

Activations Density 0.751%

phrases related to events or conferences and organizational names

occurrences of geographical locations and events

No Comments

No Known Activations