INDEX

Explanations

references to cloning or duplicating

oai_token-act-pair · gpt-3.5-turbo

repetitions of the word "clone" in various contexts

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

ourning

-0.88

akh

-0.83

akening

-0.80

erred

-0.78

orney

-0.78

alties

-0.75

tri

-0.74

GH

-0.73

Population

-0.71

ashington

-0.70

POSITIVE LOGITS

 trooper

0.95

 clones

0.88

 clone

0.88

 troopers

0.81

 Clone

0.76

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.74

 Trooper

0.71

etically

0.70

 Yamato

0.69

oning

0.66

Activations Density 0.023%

references to cloning or duplicating

repetitions of the word "clone" in various contexts

No Comments

No Known Activations