INDEX

Explanations

mentions of the word "Med" or "Medellin"

oai_token-act-pair · gpt-3.5-turbo

references to a specific medical or healthcare entity, particularly involving the term "Med."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

âĻ¦

-0.77

âĢ¢âĢ¢

-0.75

Reloaded

-0.69

cipline

-0.69

=-=-=-=-=-=-=-=-

-0.69

AUT

-0.68

 Orwell

-0.68

 innocence

-0.68

ãģį

-0.67

 Wilde

-0.67

POSITIVE LOGITS

ications

1.13

iator

0.99

undai

0.97

iation

0.97

Med

0.97

iterranean

0.95

itations

0.94

allion

0.94

Med

0.89

istrate

0.88

Activations Density 0.005%

mentions of the word "Med" or "Medellin"

references to a specific medical or healthcare entity, particularly involving the term "Med."

No Comments

No Known Activations