INDEX

Explanations

the word "our" with a high activation value

oai_token-act-pair · gpt-3.5-turbo

possessive pronouns indicating ownership or belonging

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 silent

-0.61

 glitch

-0.60

 missing

-0.57

haz

-0.56

nod

-0.56

 crossover

-0.56

 undead

-0.55

 intellig

-0.55

 losers

-0.55

 loser

-0.55

POSITIVE LOGITS

our

4.54

ours

2.97

ouring

2.61

oured

2.54

OUR

2.54

orous

1.46

orously

1.41

ourn

1.37

ourage

1.29

bour

1.27

Activations Density 0.018%

the word "our" with a high activation value

possessive pronouns indicating ownership or belonging

No Comments

No Known Activations