INDEX

Explanations

keywords related to social media

oai_token-act-pair · gpt-3.5-turbo

references to social media

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iHUD

-0.75

bestos

-0.71

ellectual

-0.70

ŃĶ

-0.70

lein

-0.67

uman

-0.66

[|

-0.66

lectic

-0.64

olphin

-0.63

 Forsaken

-0.62

POSITIVE LOGITS

 platforms

1.37

 postings

1.16

 hasht

1.11

 outlets

1.10

 posts

1.10

 sites

1.08

 networks

1.07

 accounts

1.07

 profiles

1.05

 savvy

1.04

Activations Density 0.064%

keywords related to social media

references to social media

No Comments

No Known Activations