INDEX

Explanations

references to historical context or background

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

hap

-0.74

wered

-0.74

ando

-0.73

ower

-0.69

xe

-0.68

 peas

-0.68

emouth

-0.67

wagen

-0.66

amina

-0.65

agers

-0.65

POSITIVE LOGITS

 dating

0.83

 stretching

0.77

acqu

0.77

 associ

0.77

 precedent

0.74

 affiliation

0.74

 documented

0.72

 brewing

0.70

 chronic

0.70

 history

0.69

Activations Density 0.031%

references to historical context or background

No Comments

No Known Activations