INDEX

Explanations

the word "ought."

oai_token-act-pair · gpt-3.5-turbo

expressions of obligation or advice

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

Arab

-0.66

omics

-0.63

 Polo

-0.60

packs

-0.59

dream

-0.59

ula

-0.59

ZI

-0.59

 Puzzle

-0.59

oric

-0.58

gren

-0.57

POSITIVE LOGITS

EStream

0.86

lessly

0.84

 disclaim

0.76

ĺħ

0.74

ratulations

0.74

rils

0.73

entious

0.73

DragonMagazine

0.73

ILY

0.71

lest

0.71

Activations Density 0.007%

the word "ought."

expressions of obligation or advice

No Comments

No Known Activations