INDEX

Explanations

phrases related to specific geographic locations

oai_token-act-pair · gpt-3.5-turbo

references to specific locations or geographic entities

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

 dissu

-0.72

 wanting

-0.66

 wherever

-0.66

 sacrificing

-0.65

 endeav

-0.64

 bashing

-0.62

xual

-0.62

 Cyborg

-0.61

 everywhere

-0.61

cules

-0.60

POSITIVE LOGITS

ramid

0.94

WASHINGTON

0.89

SHARE

0.85

PORT

0.85

arter

0.81

ÂłÂłÂł

0.80

ÂłÂłÂłÂłÂłÂłÂłÂł

0.79

Posted

0.79

PROV

0.78

Filename

0.78

Activations Density 0.334%

phrases related to specific geographic locations

references to specific locations or geographic entities

No Comments

No Known Activations