INDEX

Explanations

statements related to fact-checking and verification

oai_token-act-pair · gpt-3.5-turbo

phrases and terms related to fact-checking and verification processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Receiver

-0.78

 Passenger

-0.71

 arrivals

-0.65

Ell

-0.65

Pas

-0.65

 Seasons

-0.64

 docks

-0.64

 exits

-0.63

 opener

-0.63

 Manz

-0.62

POSITIVE LOGITS

abiding

1.41

based

1.39

driven

1.36

loving

1.31

breaking

1.30

enforcement

1.24

laden

1.24

oriented

1.23

bending

1.22

seeking

1.19

Activations Density 0.049%

statements related to fact-checking and verification

phrases and terms related to fact-checking and verification processes

No Comments

No Known Activations