INDEX

Explanations

references to specific parts or components of an object

oai_token-act-pair · gpt-3.5-turbo

references to components or parts of objects

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

 Predators

-0.82

 blockers

-0.62

elsius

-0.60

 raids

-0.60

âĸ¬

-0.59

quickShipAvailable

-0.59

iant

-0.57

 footh

-0.57

 blat

-0.55

qus

-0.55

POSITIVE LOGITS

meal

1.79

 piece

0.85

holder

0.85

work

0.84

ridges

0.82

glass

0.82

horn

0.77

 Piece

0.75

parts

0.75

wise

0.74

Activations Density 0.021%

references to specific parts or components of an object

references to components or parts of objects

No Comments

No Known Activations