INDEX

Explanations

references to popular movie titles, particularly superhero and action films

oai_token-act-pair · gpt-3.5-turbo

references to the Guardians of the Galaxy franchise and its associated films

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

hus

-0.81

obar

-0.81

minist

-0.76

opathic

-0.73

 administ

-0.72

grain

-0.71

 Blackwell

-0.69

istically

-0.69

ifling

-0.69

yre

-0.67

POSITIVE LOGITS

 Awakens

0.86

 Franchise

0.84

buster

0.83

 sequels

0.82

 Turtles

0.82

 villains

0.80

 Andromeda

0.78

 movies

0.76

naire

0.74

 Transformers

0.74

Activations Density 0.042%

references to popular movie titles, particularly superhero and action films

references to the Guardians of the Galaxy franchise and its associated films

No Comments

No Known Activations