INDEX

Explanations

terms related to medical conditions, particularly diseases or disorders ending in 'ph'

oai_token-act-pair · gpt-3.5-turbo

terms related to physical conditions or disorders

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

éĹĺ

-0.81

BIL

-0.78

ãģ®éŃĶ

-0.77

 Progressive

-0.74

vers

-0.71

Paw

-0.70

ELY

-0.70

ãĥ´

-0.67

cmp

-0.65

 Pace

-0.64

POSITIVE LOGITS

osate

1.53

ysics

1.32

antom

1.31

osph

1.31

obia

1.23

tml

1.15

oenix

1.05

inx

1.04

ysical

1.03

onso

1.02

Activations Density 0.019%

terms related to medical conditions, particularly diseases or disorders ending in 'ph'

terms related to physical conditions or disorders

No Comments

No Known Activations