INDEX

Explanations

references to political figures and their actions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Alic

-0.72

 Loki

-0.68

Aph

-0.66

 Spectre

-0.66

 Prism

-0.65

 Reincarnated

-0.65

 Philips

-0.64

 âĤ¬

-0.64

 Lego

-0.63

rists

-0.63

POSITIVE LOGITS

WASHINGTON

0.82

ÂłÂłÂłÂł

0.79

CLOSE

0.79

ÂłÂłÂł

0.79

Correct

0.78

Washington

0.76

pmwiki

0.76

LOS

0.75

Correction

0.73

SHARE

0.72

Activations Density 0.078%

references to political figures and their actions

No Comments

No Known Activations