INDEX

Explanations

capitalized words related to different entities

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names, places, or titles associated with specific subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ãĥīãĥ©ãĤ´ãĥ³

-0.69

ayed

-0.69

ulner

-0.69

internal

-0.68

sembly

-0.66

acting

-0.65

esp

-0.64

jri

-0.63

 distingu

-0.63

ãģĹ

-0.62

POSITIVE LOGITS

 Studio

0.73

 Collective

0.72

 Shop

0.70

 Labs

0.70

 Friends

0.69

 Alloy

0.67

 Brothers

0.66

 strives

0.66

pedia

0.66

 Flight

0.64

Activations Density 0.335%

capitalized words related to different entities

proper nouns, particularly names, places, or titles associated with specific subjects

No Comments

No Known Activations