INDEX

Explanations

numerical values, especially those related to measurements and counts

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

Reviewer

-0.93

andi

-0.82

lance

-0.80

NAS

-0.80

BALL

-0.75

tag

-0.70

alogy

-0.69

MSN

-0.69

acus

-0.68

Joy

-0.67

POSITIVE LOGITS

 million

0.90

 inches

0.88

 percent

0.87

 times

0.85

 thousand

0.84

 thirds

0.84

 hundred

0.82

 inch

0.78

 weeks

0.77

 cycles

0.76

Activations Density 0.032%

numerical values, especially those related to measurements and counts

No Comments

No Known Activations