INDEX

Explanations

zip codes in the text

oai_token-act-pair · gpt-3.5-turbo 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

raising

-0.81

ership

-0.78

ertodd

-0.77

emies

-0.76

 Democr

-0.75

ufact

-0.75

nces

-0.74

igans

-0.74

Â¯Â¯

-0.73

amental

-0.72

POSITIVE LOGITS

Zip

0.83

zip

0.81

INO

0.76

PER

0.76

Zip

0.74

ijn

0.74

sum

0.73

!/

0.70

pees

0.69

codes

0.68

Activations Density 0.014%

zip codes in the text

No Comments

No Known Activations