INDEX

Explanations

references to cages

oai_token-act-pair · gpt-3.5-turbo

references to cages

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

 Bundes

-0.73

ACTED

-0.69

 Nare

-0.68

merce

-0.66

 lender

-0.64

ibel

-0.64

IELD

-0.62

isance

-0.62

igate

-0.60

nee

-0.60

POSITIVE LOGITS

 cage

0.93

 Cage

0.92

 cages

0.87

mong

0.87

pit

0.85

washer

0.84

ModLoader

0.84

door

0.80

hold

0.77

lov

0.74

Activations Density 0.023%

references to cages

references to cages

No Comments

No Known Activations