INDEX

Explanations

references to "bots" or similar terms, potentially in the context of technology or automation

oai_token-act-pair · gpt-3.5-turbo

references to 'bots' or robotic entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ŃĶ

-0.68

 intimid

-0.67

 witnesses

-0.66

 warranties

-0.66

 disclosure

-0.66

 streng

-0.65

 prosec

-0.63

 prosecut

-0.62

 experien

-0.61

awa

-0.61

POSITIVE LOGITS

wana

1.32

anu

1.09

tery

1.03

heet

1.03

 Aires

1.01

assium

1.00

aurus

0.99

nikov

0.94

uba

0.94

weet

0.92

Activations Density 0.013%

references to "bots" or similar terms, potentially in the context of technology or automation

references to 'bots' or robotic entities

No Comments

No Known Activations