INDEX

Explanations

The neuron is looking for chemical compounds containing the word "nit"

oai_token-act-pair · gpt-3.5-turbo

references to nitrogen compounds and their effects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

xual

-0.88

ngth

-0.87

DragonMagazine

-0.86

NetMessage

-0.79

OHN

-0.75

EStream

-0.72

hip

-0.71

Haunted

-0.71

adena

-0.68

atchewan

-0.68

POSITIVE LOGITS

rous

1.12

picking

1.10

rogen

0.97

nit

0.93

rates

0.89

pick

0.81

ritional

0.81

Nit

0.81

nit

0.80

rite

0.80

Activations Density 0.013%

The neuron is looking for chemical compounds containing the word "nit"

references to nitrogen compounds and their effects

No Comments

No Known Activations