INDEX

Explanations

occurrences of the word “and” in various contexts

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

meet

-0.75

ummer

-0.75

iors

-0.71

ardy

-0.70

Leaks

-0.70

ationally

-0.70

lems

-0.69

iev

-0.68

>[

-0.67

humans

-0.67

POSITIVE LOGITS

 thence

0.92

 thereby

0.89

 consequ

0.85

 assorted

0.83

 thus

0.81

 accompanying

0.81

 consequently

0.80

 optionally

0.80

 bracelet

0.80

 hence

0.79

Activations Density 0.173%

occurrences of the word “and” in various contexts

No Comments

No Known Activations