INDEX

Explanations

medical or physical conditions and terms related to healthcare

oai_token-act-pair · gpt-3.5-turbo

references to medical conditions and injuries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Its

-0.79

enhagen

-0.69

nesday

-0.67

Its

-0.65

 Needs

-0.65

 Alright

-0.63

mie

-0.62

-->

-0.62

atin

-0.62

ellen

-0.61

POSITIVE LOGITS

 histories

0.97

 unsus

0.90

 innocuous

0.89

 dubious

0.89

 spouses

0.88

 elaborate

0.88

 fleeting

0.87

 weddings

0.87

 fortunes

0.87

 miscar

0.87

Activations Density 0.815%

medical or physical conditions and terms related to healthcare

references to medical conditions and injuries

No Comments

No Known Activations