INDEX

Explanations

Twitter handles, likely focused on promoting individuals or businesses

oai_token-act-pair · gpt-3.5-turbo

references to social media, specifically Twitter mentions and handles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

uchin

-0.68

erella

-0.67

itution

-0.65

adr

-0.63

pread

-0.63

otypes

-0.63

avorite

-0.63

pecting

-0.63

tons

-0.62

 Reincarn

-0.61

POSITIVE LOGITS

NOW

0.74

 Username

0.69

(@

0.68

":["

0.67

 incre

0.64

 HERE

0.64

Activity

0.62

 Topic

0.61

 intermitt

0.61

upd

0.61

Activations Density 0.042%

Twitter handles, likely focused on promoting individuals or businesses

references to social media, specifically Twitter mentions and handles

No Comments

No Known Activations