INDEX

Explanations

terms related to medical conditions and treatments, particularly seizures and epilepsy

oai_token-act-pair · gpt-3.5-turbo

references to seizures and related medical conditions

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

conn

-0.74

 Architects

-0.63

hidden

-0.62

 Humanity

-0.61

 Construct

-0.61

 Polit

-0.60

nan

-0.60

Kin

-0.58

 Brig

-0.58

 Crimson

-0.57

POSITIVE LOGITS

 seizure

1.44

 seizures

1.39

 epile

1.04

 epilepsy

1.00

unia

0.82

ota

0.77

ilage

0.76

inson

0.71

GBT

0.67

 confiscated

0.67

Activations Density 0.007%

terms related to medical conditions and treatments, particularly seizures and epilepsy

references to seizures and related medical conditions

No Comments

No Known Activations