INDEX

Explanations

phrases indicating potential danger or vulnerability

oai_token-act-pair · gpt-3.5-turbo

phrases related to being at risk or in danger

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

elf

-0.69

ovy

-0.66

anche

-0.64

chet

-0.63

nice

-0.62

olon

-0.60

MX

-0.60

ILA

-0.60

tein

-0.59

çĦ

-0.59

POSITIVE LOGITS

 financially

0.91

 endanger

0.83

angering

0.78

istar

0.73

ħĭ

0.72

 undermin

0.71

 lest

0.71

 posed

0.70

 mentally

0.69

alert

0.67

Activations Density 0.050%

phrases indicating potential danger or vulnerability

phrases related to being at risk or in danger

No Comments

No Known Activations