INDEX

Explanations

references to the Associated Press (AP) in news contexts

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

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_slefr-ajt/2-res_slefr-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Trinidad

-0.69

 Brand

-0.67

 Villa

-0.66

Ca

-0.65

 Berm

-0.65

 Sands

-0.64

Lod

-0.64

 Monarch

-0.63

 Ridley

-0.63

 Banks

-0.63

POSITIVE LOGITS

PLIED

1.08

PLIC

1.05

artment

0.94

AP

0.91

rompt

0.90

ocry

0.89

ACK

0.89

ILY

0.88

OE

0.88

insula

0.88

Activations Density 0.048%

references to the Associated Press (AP) in news contexts

No Comments

No Known Activations