INDEX

Explanations

phrases or sentences mentioning accusations or allegations against someone or something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "accused" indicating claims of wrongdoing or allegations

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

emetery

-0.91

ARCH

-0.81

 tremend

-0.71

ental

-0.68

lot

-0.68

[|

-0.68

arp

-0.67

erved

-0.65

hare

-0.65

OPLE

-0.64

POSITIVE LOGITS

 accuse

0.95

 accusing

0.86

 accuses

0.84

ologists

0.79

 accused

0.77

ologist

0.77

iary

0.75

onia

0.75

ologies

0.75

isance

0.74

Activations Density 0.026%

phrases or sentences mentioning accusations or allegations against someone or something

instances of the word "accused" indicating claims of wrongdoing or allegations

No Comments

No Known Activations