INDEX

Explanations

individuals or entities, particularly focusing on their importance or role within a context

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

olves

-0.74

acebook

-0.70

 exist

-0.61

units

-0.61

interstitial

-0.60

¥µ

-0.60

mint

-0.60

 upkeep

-0.60

 imaginable

-0.59

 outputs

-0.59

POSITIVE LOGITS

 meanwhile

1.09

 supra

0.99

LLP

0.89

nee

0.89

et

0.83

 however

0.73

Ore

0.72

enegger

0.72

 citing

0.70

 Associates

0.69

Activations Density 0.076%

individuals or entities, particularly focusing on their importance or role within a context

No Comments

No Known Activations