INDEX

Explanations

occurrences of ordinal numbers referring to positions or steps in a sequence

oai_token-act-pair · gpt-3.5-turbo

terms related to sequential order or chronological events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Twins

-0.70

ottest

-0.65

asus

-0.64

assies

-0.64

Items

-0.63

 Isles

-0.62

 Wand

-0.62

 Tant

-0.62

 Organizations

-0.60

enery

-0.60

POSITIVE LOGITS

 approximation

0.80

0.77

stroke

0.71

âĢĲ

0.70

 impression

0.67

 baseman

0.67

sale

0.66

mortem

0.64

 apiece

0.64

knit

0.63

Activations Density 0.125%

occurrences of ordinal numbers referring to positions or steps in a sequence

terms related to sequential order or chronological events

No Comments

No Known Activations