INDEX

Explanations

specific terms and concepts related to analysis and evaluation

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

 Helpful

-0.69

 Â®

-0.61

 Converted

-0.59

vernment

-0.57

 Sporting

-0.56

 QuÃ©

-0.55

Dou

-0.55

 Flavoring

-0.54

 DARK

-0.53

 stray

-0.52

POSITIVE LOGITS

classes

0.90

share

0.88

book

0.86

code

0.86

frame

0.82

piece

0.81

fleet

0.81

set

0.80

group

0.79

sheet

0.78

Activations Density 0.479%

specific terms and concepts related to analysis and evaluation

No Comments

No Known Activations