INDEX

Explanations

phrases related to things being unhinged or unhinging

oai_token-act-pair · gpt-3.5-turbo

words related to a state of being exposed or vulnerable

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

classified

-0.67

ACTED

-0.66

 servicing

-0.62

 forgiveness

-0.61

 forgiven

-0.60

 Sunny

-0.59

ascade

-0.59

brand

-0.57

isible

-0.57

cci

-0.57

POSITIVE LOGITS

tons

1.36

ton

0.97

\\\\\\\\

0.91

redients

0.86

leys

0.86

NESS

0.85

rina

0.85

ham

0.79

thing

0.78

eers

0.76

Activations Density 0.008%

phrases related to things being unhinged or unhinging

words related to a state of being exposed or vulnerable

No Comments

No Known Activations