INDEX

Explanations

phrases indicating the movement of people or objects, typically in relation to illegal activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

QUI

-0.92

natureconservancy

-0.73

Extra

-0.69

sure

-0.67

 teasp

-0.66

 è£ıè

-0.65

guards

-0.65

Statement

-0.65

cast

-0.64

Enlarge

-0.64

POSITIVE LOGITS

 whence

0.99

 afar

0.99

 thence

0.97

 inception

0.82

 onward

0.72

 obscurity

0.72

 Budapest

0.71

 across

0.71

 Buenos

0.70

 Tacoma

0.70

Activations Density 0.077%

phrases indicating the movement of people or objects, typically in relation to illegal activities

No Comments

No Known Activations