INDEX

Explanations

mentions of "sap"

oai_token-act-pair · gpt-3.5-turbo

references to sap

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

eering

-0.86

ated

-0.78

ãĥĦ

-0.75

abel

-0.70

naires

-0.69

ships

-0.66

ating

-0.64

eer

-0.63

 Camel

-0.63

++++++++++++++++

-0.63

POSITIVE LOGITS

sap

1.42

Sap

1.03

ient

0.94

ience

0.94

posium

0.91

olicy

0.91

iens

0.89

hetics

0.89

herer

0.85

pered

0.84

Activations Density 0.016%

mentions of "sap"

references to sap

No Comments

No Known Activations