INDEX

Explanations

mentions of individuals being hospitalized

oai_token-act-pair · gpt-3.5-turbo

terms related to hospitalization or being unable to function due to injury or illness

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

hig

-0.74

ies

-0.69

$$$$

-0.66

zeb

-0.65

profits

-0.65

axe

-0.64

neg

-0.63

 CLSID

-0.62

estamp

-0.62

DN

-0.61

POSITIVE LOGITS

 hospitalized

0.98

 schizophren

0.75

 Patients

0.74

estinal

0.71

 Nikola

0.68

inois

0.67

icter

0.67

 diagnosed

0.65

 patients

0.65

iment

0.65

Activations Density 0.013%

mentions of individuals being hospitalized

terms related to hospitalization or being unable to function due to injury or illness

No Comments

No Known Activations