INDEX

Explanations

references to anonymity, especially related to anonymous tips or communication

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Anonymous" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

eele

-0.83

enegger

-0.78

=-=-=-=-=-=-=-=-

-0.78

++++++++++++++++

-0.71

meaning

-0.71

 Gork

-0.70

asters

-0.66

efully

-0.65

fare

-0.65

asting

-0.64

POSITIVE LOGITS

 Anonymous

0.80

 obliged

0.76

TERN

0.72

iate

0.70

Anonymous

0.70

 hacker

0.68

cott

0.67

ividual

0.66

imedia

0.66

ica

0.65

Activations Density 0.007%

references to anonymity, especially related to anonymous tips or communication

occurrences of the term "Anonymous" and related terms

No Comments

No Known Activations