INDEX

Explanations

proper nouns and technical terms, potentially related to acquisitions and publications

oai_token-act-pair · gpt-3.5-turbo

references to hotels and accommodations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ollywood

-0.72

illance

-0.70

TAIN

-0.68

enegger

-0.67

NFL

-0.67

nsic

-0.67

 media

-0.65

 entertainment

-0.64

itbart

-0.63

 dressing

-0.63

POSITIVE LOGITS

 Fract

0.93

 Carib

0.91

 Lumin

0.91

 Dign

0.90

 Meadow

0.90

 Pillar

0.89

 Verd

0.86

Wid

0.85

Ori

0.84

Alm

0.84

Activations Density 0.860%

proper nouns and technical terms, potentially related to acquisitions and publications

references to hotels and accommodations

No Comments

No Known Activations