INDEX

Explanations

mentions of specific terms, potentially related to ads or sponsorship

oai_token-act-pair · gpt-3.5-turbo

mentions of "AD" related to defense systems or military contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ãĤ¡

-0.81

 pains

-0.73

 shorth

-0.70

ĸļ

-0.65

 Osiris

-0.64

 Bolton

-0.62

ãĤ©

-0.62

Ń·

-0.62

 sidx

-0.61

 tremend

-0.59

POSITIVE LOGITS

venture

1.22

VERTIS

1.14

AD

1.13

vance

1.08

VERTISEMENT

1.03

ventures

1.02

SPA

1.02

HCR

0.99

olph

0.97

vertis

0.95

Activations Density 0.007%

mentions of specific terms, potentially related to ads or sponsorship

mentions of "AD" related to defense systems or military contexts

No Comments

No Known Activations