INDEX

Explanations

phrases related to a specific medical condition code (EE) and potentially associated symptoms

oai_token-act-pair · gpt-3.5-turbo

references to specific patients or subjects in a medical or clinical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

otine

-0.91

ique

-0.81

ifiers

-0.79

Reply

-0.73

ifier

-0.73

stroke

-0.69

match

-0.69

mort

-0.68

iques

-0.68

 Kind

-0.67

POSITIVE LOGITS

EEE

1.15

EE

1.08

TY

1.08

KS

1.05

LECT

1.04

FI

0.97

KN

0.97

ZE

0.95

VEL

0.94

ASED

0.92

Activations Density 0.005%

phrases related to a specific medical condition code (EE) and potentially associated symptoms

references to specific patients or subjects in a medical or clinical context

No Comments

No Known Activations