INDEX

Explanations

words relating to actions or states of continuation or change

oai_token-act-pair · gpt-3.5-turbo

words indicating ongoing or future actions and reiterations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.80

 Sets

-0.67

GGGGGGGG

-0.65

CLASSIFIED

-0.63

quickShipAvailable

-0.62

Awesome

-0.62

ONSORED

-0.61

ufact

-0.57

 Simulator

-0.56

VERTISEMENT

-0.56

POSITIVE LOGITS

be

1.47

 recognise

1.06

 tolerate

1.05

 serve

1.04

 allow

1.04

 benefit

1.01

 require

1.01

 suffice

1.01

 succumb

1.01

 qualify

1.00

Activations Density 0.158%

words relating to actions or states of continuation or change

words indicating ongoing or future actions and reiterations

No Comments

No Known Activations