INDEX

Explanations

verbs related to action, often in the form of verbs ending in 's' or 'es'

oai_token-act-pair · gpt-3.5-turbo

action-oriented verbs that imply evaluation, adjustment, or significant consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

lug

-0.71

 stocking

-0.62

 acknow

-0.62

tyr

-0.61

kin

-0.60

 embark

-0.59

 Guer

-0.56

 parting

-0.56

 aven

-0.56

 princ

-0.56

POSITIVE LOGITS

es

1.41

ifies

1.39

ulates

1.35

eth

1.33

uates

1.32

ends

1.32

ates

1.31

uces

1.26

inates

1.24

zes

1.20

Activations Density 0.188%

verbs related to action, often in the form of verbs ending in 's' or 'es'

action-oriented verbs that imply evaluation, adjustment, or significant consequences

No Comments

No Known Activations