INDEX

Explanations

numerical values and measurements related to cooking times and serving sizes

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

pse

-0.84

 folklore

-0.67

edu

-0.66

 Ples

-0.63

 discredited

-0.62

 corruption

-0.62

 bureaucr

-0.62

 exploits

-0.61

une

-0.60

æ©Ł

-0.60

POSITIVE LOGITS

 servings

1.28

Ingredients

0.98

 Ingredients

0.94

Includes

0.90

 Calories

0.90

66666666

0.79

 Serving

0.78

packs

0.77

lbs

0.76

rooms

0.75

Activations Density 0.083%

numerical values and measurements related to cooking times and serving sizes

No Comments

No Known Activations