INDEX

Explanations

phrases related to involvement in various activities or incidents

oai_token-act-pair · gpt-3.5-turbo

the phrase "involved in" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.81

Reviewed

-0.76

ratulations

-0.74

uce

-0.73

 entrants

-0.69

edu

-0.66

rious

-0.65

ious

-0.60

 afraid

-0.60

 emerges

-0.60

POSITIVE LOGITS

ordinate

0.90

 prostitution

0.71

pport

0.69

roads

0.68

 behalf

0.66

 politics

0.66

acas

0.66

 accordance

0.66

organic

0.65

bound

0.65

Activations Density 0.081%

phrases related to involvement in various activities or incidents

the phrase "involved in" and its variations

No Comments

No Known Activations