INDEX

Explanations

phrases related to medical professionals or experts

oai_token-act-pair · gpt-3.5-turbo

references to medical professionals or doctors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 actionGroup

-0.83

 chorus

-0.76

LOAD

-0.66

CHAT

-0.65

 potatoes

-0.65

RANT

-0.65

HUD

-0.64

SHIP

-0.63

OPA

-0.63

 Territories

-0.63

POSITIVE LOGITS

umin

1.11

inker

1.04

inks

1.00

inking

0.99

illing

0.98

udge

0.98

agnar

0.96

ifts

0.96

ink

0.94

ifting

0.93

Activations Density 0.014%

phrases related to medical professionals or experts

references to medical professionals or doctors

No Comments

No Known Activations