INDEX

Explanations

instances of numerical quantities followed by a sequence of specific keywords

oai_token-act-pair · gpt-3.5-turbo

references to quantities of two or four

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.4.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

NOT

-0.65

 Soldier

-0.63

 Thor

-0.62

patient

-0.62

Thor

-0.62

adder

-0.62

 Materials

-0.61

 Warfare

-0.61

ller

-0.61

 Â¶

-0.61

POSITIVE LOGITS

 olds

0.73

undred

0.73

pei

0.72

ilk

0.71

irty

0.68

endif

0.68

 versa

0.66

igree

0.66

enged

0.65

luent

0.65

Activations Density 0.095%

instances of numerical quantities followed by a sequence of specific keywords

references to quantities of two or four

No Comments

No Known Activations