INDEX

Explanations

locations or places mentioned in text

oai_token-act-pair · gpt-3.5-turbo

geographical locations or specific place names

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

ictions

-0.97

iction

-0.86

 sentence

-0.78

xon

-0.71

employed

-0.71

 clutch

-0.71

ersive

-0.71

*/(

-0.69

apt

-0.69

anguage

-0.68

POSITIVE LOGITS

 Sina

0.88

 Augusta

0.83

 Camp

0.83

 Gibraltar

0.82

 Canaveral

0.82

Kat

0.80

 Tours

0.80

 Reconstruction

0.77

 Royale

0.77

 George

0.77

Activations Density 0.193%

locations or places mentioned in text

geographical locations or specific place names

No Comments

No Known Activations