INDEX

Explanations

specific words related to positive outcomes or solutions

oai_token-act-pair · gpt-3.5-turbo

terms related to causes and benefits of various issues or concepts

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

ALLY

-0.89

zzi

-0.72

lette

-0.72

LET

-0.72

ski

-0.66

DonaldTrump

-0.66

END

-0.64

ANA

-0.64

Street

-0.63

psc

-0.63

POSITIVE LOGITS

etter

1.06

etting

1.00

pring

0.98

 surrounding

0.87

afety

0.86

cape

0.85

mith

0.85

 inherent

0.79

ettings

0.79

omething

0.78

Activations Density 0.273%

specific words related to positive outcomes or solutions

terms related to causes and benefits of various issues or concepts

No Comments

No Known Activations