INDEX

Explanations

titles of movies, TV shows, and other media

oai_token-act-pair · gpt-3.5-turbo

quoted text and titles of media works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ishers

-0.80

roleum

-0.72

ioned

-0.70

amination

-0.69

ons

-0.68

lying

-0.67

emonium

-0.65

amiliar

-0.65

icons

-0.65

iology

-0.65

POSITIVE LOGITS

 Remastered

1.01

 soundtrack

0.96

 starring

0.94

 lyrics

0.91

 premiered

0.88

 trilogy

0.87

 premie

0.86

 finale

0.85

 Season

0.82

 sequel

0.81

Activations Density 0.232%

titles of movies, TV shows, and other media

quoted text and titles of media works

No Comments

No Known Activations