INDEX

Explanations

words related to health conditions and medical terms

oai_token-act-pair · gpt-3.5-turbo

terminology and concepts related to medical conditions and their implications

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

 guiActiveUnfocused

-0.65

 autonom

-0.53

 respectively

-0.52

 Ibid

-0.51

 converge

-0.50

 expire

-0.50

caf

-0.50

 delinquent

-0.50

 captive

-0.49

 hereafter

-0.49

POSITIVE LOGITS

ī

0.86

hig

0.86

ActionCode

0.83

Ă

0.83

 RandomRedditor

0.83

þ

0.83

 TheNitrome

0.83

ü

0.82

Ą

0.82

Activations Density 1.635%

words related to health conditions and medical terms

terminology and concepts related to medical conditions and their implications

No Comments

No Known Activations