INDEX

Explanations

spoiler warnings in texts

oai_token-act-pair · gpt-3.5-turbo

references to spoilers in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

llan

-0.73

trak

-0.72

orney

-0.71

undai

-0.70

kes

-0.69

 Architects

-0.68

 Palest

-0.67

kos

-0.67

Motion

-0.67

urat

-0.66

POSITIVE LOGITS

 spoiler

0.95

 spoilers

0.94

OIL

0.92

OUS

0.84

 spoil

0.82

ervative

0.75

spo

0.74

ific

0.73

 alert

0.73

pedia

0.72

Activations Density 0.077%

spoiler warnings in texts

references to spoilers in the text

No Comments

No Known Activations