INDEX

Explanations

adverbs modifying the competency or intensity of an action

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a degree of certainty or approximation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

agate

-0.86

ourses

-0.69

Trend

-0.68

eria

-0.66

EMENT

-0.65

alf

-0.64

orrow

-0.63

 Provision

-0.63

aley

-0.62

illary

-0.62

POSITIVE LOGITS

 unint

0.81

earchers

0.73

 mids

0.72

 exclusively

0.71

 identical

0.70

untarily

0.70

 overpower

0.67

 unheard

0.66

ali

0.66

 indist

0.66

Activations Density 0.025%

adverbs modifying the competency or intensity of an action

phrases indicating a degree of certainty or approximation

No Comments

No Known Activations