INDEX

Explanations

instances or mentions of significant events or disruptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

¶ħ

-0.67

WP

-0.63

orer

-0.61

ĩ

-0.60

 âĢķ

-0.57

ORD

-0.57

igl

-0.56

Deploy

-0.56

ãĢ

-0.55

USS

-0.54

POSITIVE LOGITS

 respectively

1.06

 albeit

0.96

etc

0.80

uh

0.61

 according

0.60

otos

0.59

 disg

0.59

 dunno

0.56

but

0.55

um

0.54

Activations Density 0.480%

instances or mentions of significant events or disruptions

No Comments

No Known Activations