INDEX

Explanations

individuals involved in various interactions or scenarios

oai_token-act-pair · gpt-3.5-turbo

terms related to individuals in various roles or capacities

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

 VIDEOS

-0.74

issues

-0.74

 instruments

-0.73

ernels

-0.71

 levers

-0.71

 Pieces

-0.70

 calendars

-0.70

 Dragonbound

-0.69

adders

-0.69

 bindings

-0.69

POSITIVE LOGITS

 whom

1.41

who

1.41

 named

1.28

 whose

1.25

who

1.10

whose

1.00

 surn

0.95

 nicknamed

0.92

friend

0.90

 friend

0.83

Activations Density 0.263%

individuals involved in various interactions or scenarios

terms related to individuals in various roles or capacities

No Comments

No Known Activations