INDEX

Explanations

references to a specific person named "Koster"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Koster"

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

ccording

-0.81

andro

-0.70

trak

-0.68

ĵĺ

-0.65

 ILCS

-0.65

Ĥª

-0.64

%]

-0.62

 princ

-0.62

kai

-0.61

£ı

-0.61

POSITIVE LOGITS

oster

0.91

eers

0.77

pole

0.74

amus

0.74

hof

0.73

velt

0.72

FORE

0.70

 Volks

0.70

osity

0.69

thal

0.69

Activations Density 0.025%

references to a specific person named "Koster"

occurrences of the word "Koster"

No Comments

No Known Activations